feat(extractors): add 17 job source extractors and cross-source dedup

Adds extractor packages: arbeitnow, ashby, careerjet, fourdayweek,
greenhouse, himalayas, jobicy, jooble, lever, reed, remoteok, remotive,
themuse, usajobs, weworkremotely, workday — each with manifest, package
metadata and README.

Pipeline / shared:
- shared/job-fingerprint: stable hash for cross-source dedup, with tests
- discover-jobs: dedup via fingerprint and richer per-source merging
- jobs repository: fingerprint-aware upsert / lookup
- settings-registry, settings types/routes, demo-defaults: knobs for the
  new sources
- shared extractors index: register the new manifests
- location-support, profiles route: small fixes for the new sources

Tooling:
- scripts/smoke-extractors.ts to sanity-check each source locally
- scripts/jobber-cron-{cherepaha,dobkin}.env.example: per-host cron
  templates (CHANGEME placeholders only)
- .env.example: documented env vars for the new extractors
- .gitignore: ignore extractors/*/storage/ runtime caches (was ukvisajobs only)

Co-authored-by: Cursor <cursoragent@cursor.com>

.env.example

@@ -97,3 +97,116 @@ ADZUNA_APP_KEY=
 # =============================================================================
 # Filter for remote-only jobs (default: 0 = disabled)
 # JOBSPY_IS_REMOTE=0
+
+# =============================================================================
+# USAJOBS API (US federal jobs) - optional, US-only
+# =============================================================================
+# Register at https://developer.usajobs.gov/APIRequest/Index
+# USAJOBS requires a User-Agent that is a real contact email (per their TOS).
+# Leave unset to disable the source.
+# USAJOBS_API_KEY=
+# USAJOBS_USER_AGENT=you@example.com
+# USAJOBS_MAX_JOBS_PER_TERM=100
+
+# =============================================================================
+# Jobicy (remote jobs feed) - optional, no auth
+# =============================================================================
+# Public JSON endpoint, capped at 50 results per call.
+# JOBICY_MAX_JOBS_PER_TERM=100
+
+# =============================================================================
+# The Muse (jobs API) - optional, API key recommended
+# =============================================================================
+# https://www.themuse.com/developers/api/v2 — works without a key but is
+# heavily rate-limited. Set THEMUSE_API_KEY for higher quotas.
+# THEMUSE_API_KEY=
+# THEMUSE_MAX_JOBS_PER_TERM=100
+
+# =============================================================================
+# Jooble (aggregator API) - optional
+# =============================================================================
+# Sign up at https://jooble.org/api/about for an API key.
+# JOOBLE_API_KEY=
+# JOOBLE_MAX_JOBS_PER_TERM=100
+
+# =============================================================================
+# Careerjet (publisher API v4) - optional
+# =============================================================================
+# Register at https://www.careerjet.com/partners/api/ — declare API key + server IP(s).
+# CAREERJET_AFFID=your_api_key
+# CAREERJET_REFERER=https://your-site.com/path-to-job-search/
+# CAREERJET_USER_IP=203.0.113.1
+# Optional override for the required user_agent query param:
+# CAREERJET_USER_AGENT=Mozilla/5.0 ...
+# CAREERJET_MAX_JOBS_PER_TERM=100
+
+# =============================================================================
+# Reed.co.uk (UK jobs API) - optional, UK-only
+# =============================================================================
+# Register at https://www.reed.co.uk/developers/jobseeker for an API key.
+# REED_API_KEY=
+# REED_MAX_JOBS_PER_TERM=100
+
+# =============================================================================
+# Remote OK (remote jobs feed) - optional, no auth
+# =============================================================================
+# Public single-shot JSON feed at https://remoteok.com/api. We filter
+# client-side by your search terms (matched against position + tags).
+# Per Remote OK's TOS, link back to the original posting URLs when republishing.
+# REMOTEOK_MAX_JOBS_PER_TERM=100
+
+# =============================================================================
+# Remotive (remote jobs feed) - optional, no auth
+# =============================================================================
+# Public JSON API at https://remotive.com/api/remote-jobs?limit=N&search=term.
+# Each search term is sent as the `search` parameter.
+# REMOTIVE_MAX_JOBS_PER_TERM=100
+
+# =============================================================================
+# Arbeitnow (multi-ATS aggregator) - optional, no auth
+# =============================================================================
+# Public JSON API at https://www.arbeitnow.com/api/job-board-api?page=N.
+# Aggregates from Greenhouse, SmartRecruiters, Join, TeamTailor, Recruitee,
+# and Comeet. No server-side search; filtering is done client-side.
+# ARBEITNOW_MAX_JOBS_PER_TERM=100
+
+# =============================================================================
+# Himalayas (remote jobs feed) - optional, no auth
+# =============================================================================
+# Public JSON API at https://himalayas.app/jobs/api?limit=N&offset=M.
+# No server-side search; filtering is done client-side by title + categories.
+# HIMALAYAS_MAX_JOBS_PER_TERM=100
+
+# =============================================================================
+# We Work Remotely (RSS feed) - optional, no auth
+# =============================================================================
+# Public RSS at https://weworkremotely.com/remote-jobs.rss (all categories).
+# Single fetch; filtering is done client-side by title + skills + category.
+# WEWORKREMOTELY_MAX_JOBS_PER_TERM=100
+
+# =============================================================================
+# 4 Day Week (reduced-schedule jobs) - optional, no auth
+# =============================================================================
+# Public JSON API at https://4dayweek.io/api/jobs?page=N.
+# Paginated; filtering is done client-side by title + tech stack.
+# No job description in listings; links to 4dayweek.io for details.
+# FOURDAYWEEK_MAX_JOBS_PER_TERM=100
+
+# =============================================================================
+# Public ATS sources (Lever / Ashby / Greenhouse) - optional
+# =============================================================================
+# Comma- or newline-separated company slugs. The slug is the path segment used
+# in each provider's public job board, e.g. `lever.co/some-company` → "some-company".
+# LEVER_COMPANIES=netflix,figma
+# ASHBY_COMPANIES=ramp,linear
+# GREENHOUSE_COMPANIES=stripe,airbnb
+
+# =============================================================================
+# Workday (public career sites) - optional
+# =============================================================================
+# Newline- or comma-separated entries. Each entry is either:
+#   1) A career-site URL we'll auto-parse, e.g.
+#      https://nvidia.wd5.myworkdayjobs.com/en-US/NVIDIAExternalCareerSite
+#   2) A JSON object with explicit fields:
+#      {"company":"NVIDIA","tenantUrl":"https://nvidia.wd5.myworkdayjobs.com","tenant":"nvidia","site":"NVIDIAExternalCareerSite","locale":"en-US"}
+# WORKDAY_TENANTS=

.gitignore

@@ -15,8 +15,8 @@ docs-site/build/
 # Data directory (bind mount in Docker)
 data/
 
-# Extractor storage outputs and cached auth
+# Extractor storage outputs and cached auth (per-extractor runtime data)
-extractors/ukvisajobs/storage/
+extractors/*/storage/
 
 # OS files
 .DS_Store

extractors/arbeitnow/README.md

@@ -0,0 +1,11 @@
+# arbeitnow-extractor
+
+Pulls listings from the public [Arbeitnow API](https://www.arbeitnow.com/api/job-board-api).
+
+- No authentication required.
+- Returns 100 jobs per page; we paginate up to 5 pages (500 jobs).
+- No server-side search — we filter client-side by matching title + tags
+  against each pipeline search term.
+- Aggregates postings from Greenhouse, SmartRecruiters, Join, TeamTailor,
+  Recruitee, and Comeet.
+- Caps results per term via the `arbeitnowMaxJobsPerTerm` setting (default 100).

extractors/arbeitnow/manifest.ts

@@ -0,0 +1,172 @@
+/**
+ * Arbeitnow public job board API.
+ *
+ * https://www.arbeitnow.com/api/job-board-api?page=N
+ *
+ * No auth. Returns 100 results per page, sorted by creation date.
+ * No server-side search — we paginate and filter client-side by
+ * title + tags against each pipeline search term.
+ *
+ * Aggregates listings from Greenhouse, SmartRecruiters, Join,
+ * TeamTailor, Recruitee, and Comeet.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+const API_URL = "https://www.arbeitnow.com/api/job-board-api";
+const MAX_PAGES = 5;
+
+interface ArbeitnowJob {
+  slug?: string;
+  company_name?: string;
+  title?: string;
+  description?: string;
+  remote?: boolean;
+  url?: string;
+  tags?: string[];
+  job_types?: string[];
+  location?: string;
+  created_at?: number;
+}
+
+interface ArbeitnowResponse {
+  data?: ArbeitnowJob[];
+  links?: { next?: string | null };
+  meta?: { current_page?: number };
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed || undefined;
+}
+
+function matchesTerm(job: ArbeitnowJob, term: string): boolean {
+  const lower = term.toLowerCase();
+  if (job.title?.toLowerCase().includes(lower)) return true;
+  if (
+    Array.isArray(job.tags) &&
+    job.tags.some(
+      (t) => typeof t === "string" && t.toLowerCase().includes(lower),
+    )
+  )
+    return true;
+  return false;
+}
+
+function mapJob(raw: ArbeitnowJob): CreateJobInput | null {
+  const jobUrl = asString(raw.url);
+  if (!jobUrl) return null;
+
+  const tags = Array.isArray(raw.tags)
+    ? raw.tags.filter((t): t is string => typeof t === "string" && t.length > 0)
+    : [];
+
+  const jobTypes = Array.isArray(raw.job_types)
+    ? raw.job_types
+        .filter((t): t is string => typeof t === "string" && t.length > 0)
+        .join(", ")
+    : undefined;
+
+  const datePosted =
+    typeof raw.created_at === "number"
+      ? new Date(raw.created_at * 1000).toISOString()
+      : undefined;
+
+  return {
+    source: "arbeitnow",
+    sourceJobId: asString(raw.slug),
+    title: asString(raw.title) ?? "Unknown Title",
+    employer: asString(raw.company_name) ?? "Unknown Employer",
+    jobUrl,
+    applicationLink: jobUrl,
+    location: asString(raw.location) ?? "Unknown",
+    isRemote: raw.remote === true,
+    jobType: jobTypes || undefined,
+    datePosted,
+    jobDescription: asString(raw.description),
+    disciplines: tags.length > 0 ? tags.join(", ") : undefined,
+  };
+}
+
+async function fetchPage(page: number): Promise<ArbeitnowResponse> {
+  const url = `${API_URL}?page=${page}`;
+  const response = await fetch(url, {
+    headers: { Accept: "application/json" },
+  });
+  if (!response.ok) {
+    throw new Error(`Arbeitnow request failed with status ${response.status}`);
+  }
+  return (await response.json()) as ArbeitnowResponse;
+}
+
+export const manifest: ExtractorManifest = {
+  id: "arbeitnow",
+  displayName: "Arbeitnow",
+  providesSources: ["arbeitnow"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const maxJobs = context.settings.arbeitnowMaxJobsPerTerm
+      ? Number.parseInt(context.settings.arbeitnowMaxJobsPerTerm, 10)
+      : 100;
+
+    const terms = context.searchTerms.length > 0 ? context.searchTerms : [];
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    try {
+      for (let page = 1; page <= MAX_PAGES; page += 1) {
+        if (context.shouldCancel?.()) break;
+        if (out.length >= maxJobs * Math.max(terms.length, 1)) break;
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: 0,
+          termsTotal: 1,
+          currentUrl: `page ${page}`,
+          detail: `Arbeitnow: fetching page ${page}`,
+        });
+
+        const body = await fetchPage(page);
+        const jobs = Array.isArray(body.data) ? body.data : [];
+
+        if (jobs.length === 0) break;
+
+        for (const raw of jobs) {
+          if (terms.length > 0 && !terms.some((t) => matchesTerm(raw, t))) {
+            continue;
+          }
+          const mapped = mapJob(raw);
+          if (!mapped) continue;
+          const key = mapped.sourceJobId || mapped.jobUrl;
+          if (seen.has(key)) continue;
+          seen.add(key);
+          out.push(mapped);
+        }
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: 0,
+          termsTotal: 1,
+          currentUrl: `page ${page}`,
+          jobPagesProcessed: out.length,
+          detail: `Arbeitnow: page ${page} done (${out.length} matched so far)`,
+        });
+
+        if (!body.links?.next) break;
+      }
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: out, error: message };
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/arbeitnow/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "arbeitnow-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "Arbeitnow public job board API extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/arbeitnow/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/ashby/README.md

@@ -0,0 +1,8 @@
+# ashby-extractor
+
+Public Ashby job-board feeds via
+`GET https://api.ashbyhq.com/posting-api/job-board/{company}`.
+
+- No auth.
+- Configure target slugs via `ashbyCompanies` (comma/newline) or
+  `ASHBY_COMPANIES` env.

extractors/ashby/manifest.ts

@@ -0,0 +1,185 @@
+/**
+ * Ashby public job board API.
+ *
+ * https://developers.ashbyhq.com/reference/posting-api-job-board
+ *   GET https://api.ashbyhq.com/posting-api/job-board/{company}
+ *
+ * No auth. Each entry in `ashbyCompanies` is fetched independently.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+interface AshbyAddress {
+  postalAddress?: {
+    addressLocality?: string;
+    addressRegion?: string;
+    addressCountry?: string;
+  };
+}
+interface AshbyJob {
+  id?: string;
+  title?: string;
+  jobUrl?: string;
+  applyUrl?: string;
+  publishedAt?: string;
+  employmentType?: string;
+  isRemote?: boolean;
+  team?: string;
+  department?: string;
+  location?: string;
+  locationName?: string;
+  secondaryLocations?: Array<{ location?: string; locationName?: string }>;
+  address?: AshbyAddress;
+  descriptionPlain?: string;
+  descriptionHtml?: string;
+}
+interface AshbyResponse {
+  jobs?: AshbyJob[];
+  apiVersion?: string;
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed ? trimmed : undefined;
+}
+
+function readCompanies(raw: string | undefined): string[] {
+  if (!raw) return [];
+  try {
+    const parsed = JSON.parse(raw);
+    if (Array.isArray(parsed)) {
+      return parsed
+        .map((entry) => (typeof entry === "string" ? entry.trim() : ""))
+        .filter(Boolean);
+    }
+  } catch {
+    // fall through
+  }
+  return raw
+    .split(/[\n,;|]+/)
+    .map((entry) => entry.trim())
+    .filter(Boolean);
+}
+
+function locationFor(job: AshbyJob): string | undefined {
+  const primary =
+    asString(job.locationName) ?? asString(job.location) ?? undefined;
+  const secondary =
+    job.secondaryLocations
+      ?.map((entry) => asString(entry.locationName) ?? asString(entry.location))
+      .filter((value): value is string => Boolean(value)) ?? [];
+  const all = [primary, ...secondary].filter((value): value is string =>
+    Boolean(value),
+  );
+  return all.length > 0 ? all.join("; ") : undefined;
+}
+
+function mapJob(job: AshbyJob, company: string): CreateJobInput | null {
+  const jobUrl = asString(job.jobUrl) ?? asString(job.applyUrl);
+  if (!jobUrl) return null;
+  const employer = company
+    .split(/[-_]/)
+    .filter(Boolean)
+    .map((part) => part.charAt(0).toUpperCase() + part.slice(1))
+    .join(" ");
+  return {
+    source: "ashby",
+    sourceJobId: asString(job.id),
+    title: asString(job.title) ?? "Unknown Title",
+    employer: employer || company,
+    jobUrl,
+    applicationLink: asString(job.applyUrl) ?? jobUrl,
+    location: locationFor(job),
+    isRemote: typeof job.isRemote === "boolean" ? job.isRemote : undefined,
+    jobType: asString(job.employmentType),
+    jobFunction: asString(job.team),
+    companyIndustry: asString(job.department),
+    datePosted: asString(job.publishedAt),
+    jobDescription:
+      asString(job.descriptionPlain) ?? asString(job.descriptionHtml),
+  };
+}
+
+async function fetchCompany(company: string): Promise<AshbyJob[]> {
+  const url = `https://api.ashbyhq.com/posting-api/job-board/${encodeURIComponent(company)}`;
+  const response = await fetch(url, {
+    headers: { Accept: "application/json" },
+  });
+  if (response.status === 404) return [];
+  if (!response.ok) {
+    throw new Error(
+      `Ashby request for "${company}" failed with status ${response.status}`,
+    );
+  }
+  const body = (await response.json()) as AshbyResponse;
+  return Array.isArray(body.jobs) ? body.jobs : [];
+}
+
+export const manifest: ExtractorManifest = {
+  id: "ashby",
+  displayName: "Ashby (ATS)",
+  providesSources: ["ashby"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const companies = readCompanies(context.settings.ashbyCompanies);
+    if (companies.length === 0) {
+      return {
+        success: true,
+        jobs: [],
+        error:
+          "No Ashby companies configured. Set ASHBY_COMPANIES or the ashbyCompanies setting (comma- or newline-separated slugs).",
+      };
+    }
+
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    try {
+      for (let i = 0; i < companies.length; i += 1) {
+        if (context.shouldCancel?.()) break;
+        const company = companies[i];
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i,
+          termsTotal: companies.length,
+          currentUrl: company,
+          detail: `Ashby: ${company} (${i + 1}/${companies.length})`,
+        });
+
+        let added = 0;
+        const jobs = await fetchCompany(company);
+        for (const job of jobs) {
+          const mapped = mapJob(job, company);
+          if (!mapped) continue;
+          const key = mapped.sourceJobId || mapped.jobUrl;
+          if (seen.has(key)) continue;
+          seen.add(key);
+          out.push(mapped);
+          added += 1;
+        }
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i + 1,
+          termsTotal: companies.length,
+          currentUrl: company,
+          jobPagesProcessed: out.length,
+          detail: `Ashby: ${company} → ${added} jobs (${out.length} total)`,
+        });
+      }
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: out, error: message };
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/ashby/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "ashby-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "Ashby public ATS extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/ashby/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/careerjet/README.md

@@ -0,0 +1,12 @@
+# careerjet-extractor
+
+[Careerjet publisher API v4](https://www.careerjet.com/partners/api/) (`https://search.api.careerjet.net/v4/query`).
+
+## Required configuration
+
+- **`CAREERJET_AFFID`** — Your publisher **API key** (settings key `careerjetAffid`). Used as the Basic auth **username**; password is empty.
+- **`CAREERJET_REFERER`** — The `Referer` header Careerjet requires: the full URL of the job-search page on your registered site (e.g. `https://yoursite.com/find-jobs/`).
+- **`CAREERJET_USER_IP`** — The `user_ip` query parameter. In the [publisher dashboard](https://www.careerjet.com/partners/), add your **server’s outbound IP** (and any dev machine IP) under “Server IP address”; this value should match an allowlisted address.
+- **`CAREERJET_USER_AGENT`** (optional) — Override the default `user_agent` param if Careerjet asks for a specific string.
+
+`selectedCountry` maps to `locale_code`; the first `searchCities` token is sent as `location`. Capped per term via `careerjetMaxJobsPerTerm` (default 100). The v4 API allows up to **10** pages per query.

extractors/careerjet/manifest.ts

@@ -0,0 +1,267 @@
+/**
+ * Careerjet publisher search API (v4).
+ *
+ * https://www.careerjet.com/partners/api/
+ *   GET https://search.api.careerjet.net/v4/query
+ *
+ * Uses Basic auth (username = publisher API key, password empty). Requires a
+ * Referer header and `user_ip` / `user_agent` query params. Register your
+ * server's outbound IP(s) in the Careerjet publisher dashboard.
+ *
+ * Env: CAREERJET_AFFID (API key), CAREERJET_REFERER (job-search page URL),
+ * CAREERJET_USER_IP (must match an allowlisted IP), optional CAREERJET_USER_AGENT.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+const API_URL = "https://search.api.careerjet.net/v4/query";
+
+const DEFAULT_USER_AGENT =
+  "Mozilla/5.0 (compatible; JobOps/1.0; job-search pipeline)";
+
+interface CareerjetJob {
+  title?: string;
+  description?: string;
+  company?: string;
+  salary?: string;
+  date?: string;
+  url?: string;
+  site?: string;
+  locations?: string;
+}
+interface CareerjetResponse {
+  type?: string;
+  jobs?: CareerjetJob[];
+  hits?: number;
+  pages?: number;
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed ? trimmed : undefined;
+}
+
+function mapJob(raw: CareerjetJob): CreateJobInput | null {
+  const jobUrl = asString(raw.url);
+  if (!jobUrl) return null;
+  return {
+    source: "careerjet",
+    title: asString(raw.title) ?? "Unknown Title",
+    employer: asString(raw.company) ?? "Unknown Employer",
+    jobUrl,
+    applicationLink: jobUrl,
+    location: asString(raw.locations),
+    salary: asString(raw.salary),
+    datePosted: asString(raw.date),
+    jobDescription: asString(raw.description),
+    companyDescription: asString(raw.site),
+  };
+}
+
+function localeForCountry(country: string): string {
+  const key = country.trim().toLowerCase();
+  switch (key) {
+    case "united kingdom":
+    case "uk":
+      return "en_GB";
+    case "united states":
+    case "usa":
+    case "us":
+      return "en_US";
+    case "canada":
+      return "en_CA";
+    case "australia":
+      return "en_AU";
+    case "germany":
+      return "de_DE";
+    case "france":
+      return "fr_FR";
+    case "spain":
+      return "es_ES";
+    case "italy":
+      return "it_IT";
+    case "netherlands":
+      return "nl_NL";
+    default:
+      return "en_GB";
+  }
+}
+
+function basicAuthorizationHeader(apiKey: string): string {
+  const credentials = `${apiKey}:`;
+  const encoded = Buffer.from(credentials, "utf8").toString("base64");
+  return `Basic ${encoded}`;
+}
+
+async function fetchPage(args: {
+  apiKey: string;
+  keywords: string;
+  location?: string;
+  page: number;
+  pageSize: number;
+  localeCode: string;
+  referer: string;
+  userIp: string;
+  userAgent: string;
+}): Promise<CareerjetResponse> {
+  const url = new URL(API_URL);
+  url.searchParams.set("locale_code", args.localeCode);
+  url.searchParams.set("keywords", args.keywords);
+  if (args.location) url.searchParams.set("location", args.location);
+  url.searchParams.set("page", String(args.page));
+  url.searchParams.set("page_size", String(args.pageSize));
+  url.searchParams.set("user_ip", args.userIp);
+  url.searchParams.set("user_agent", args.userAgent);
+
+  const response = await fetch(url.toString(), {
+    headers: {
+      Accept: "application/json",
+      Authorization: basicAuthorizationHeader(args.apiKey),
+      Referer: args.referer,
+    },
+  });
+  if (!response.ok) {
+    const snippet = (await response.text()).slice(0, 200);
+    throw new Error(
+      `Careerjet request failed with status ${response.status}${snippet ? `: ${snippet}` : ""}`,
+    );
+  }
+  return (await response.json()) as CareerjetResponse;
+}
+
+export const manifest: ExtractorManifest = {
+  id: "careerjet",
+  displayName: "Careerjet",
+  providesSources: ["careerjet"],
+  requiredEnvVars: [
+    "CAREERJET_AFFID",
+    "CAREERJET_REFERER",
+    "CAREERJET_USER_IP",
+  ],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const apiKey =
+      context.settings.careerjetAffid?.trim() ||
+      process.env.CAREERJET_AFFID?.trim();
+    const referer =
+      context.settings.careerjetReferer?.trim() ||
+      process.env.CAREERJET_REFERER?.trim();
+    const userIp =
+      context.settings.careerjetUserIp?.trim() ||
+      process.env.CAREERJET_USER_IP?.trim();
+    const userAgent =
+      context.settings.careerjetUserAgent?.trim() ||
+      process.env.CAREERJET_USER_AGENT?.trim() ||
+      DEFAULT_USER_AGENT;
+
+    if (!apiKey) {
+      return {
+        success: false,
+        jobs: [],
+        error:
+          "Careerjet requires CAREERJET_AFFID (publisher API key for Basic auth).",
+      };
+    }
+    if (!referer) {
+      return {
+        success: false,
+        jobs: [],
+        error:
+          "Careerjet v4 requires CAREERJET_REFERER (the Referer URL of your job-search page, per Careerjet docs).",
+      };
+    }
+    if (!userIp) {
+      return {
+        success: false,
+        jobs: [],
+        error:
+          "Careerjet v4 requires CAREERJET_USER_IP. Use an IP you have allowlisted in the Careerjet publisher dashboard (typically your server's public egress IP).",
+      };
+    }
+
+    const maxJobsPerTerm = context.settings.careerjetMaxJobsPerTerm
+      ? Number.parseInt(context.settings.careerjetMaxJobsPerTerm, 10)
+      : 100;
+    const pageSize = 50;
+    const localeCode = localeForCountry(context.selectedCountry || "");
+
+    const terms = context.searchTerms.length > 0 ? context.searchTerms : [""];
+    const location =
+      context.settings.searchCities?.split("|")[0]?.trim() || undefined;
+
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    try {
+      for (let i = 0; i < terms.length; i += 1) {
+        if (context.shouldCancel?.()) break;
+        const term = terms[i].trim();
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i,
+          termsTotal: terms.length,
+          currentUrl: term || "(all)",
+          detail: `Careerjet: term ${i + 1}/${terms.length}`,
+        });
+
+        let collected = 0;
+        let page = 1;
+        let totalPages = Number.POSITIVE_INFINITY;
+        while (
+          collected < maxJobsPerTerm &&
+          page <= totalPages &&
+          page <= 10
+        ) {
+          if (context.shouldCancel?.()) break;
+          const body = await fetchPage({
+            apiKey,
+            keywords: term,
+            location,
+            page,
+            pageSize,
+            localeCode,
+            referer,
+            userIp,
+            userAgent,
+          });
+          if (typeof body.pages === "number") totalPages = body.pages;
+          const items = Array.isArray(body.jobs) ? body.jobs : [];
+          if (items.length === 0) break;
+          for (const raw of items) {
+            const mapped = mapJob(raw);
+            if (!mapped) continue;
+            const key = mapped.jobUrl;
+            if (seen.has(key)) continue;
+            seen.add(key);
+            out.push(mapped);
+            collected += 1;
+            if (collected >= maxJobsPerTerm) break;
+          }
+          page += 1;
+        }
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i + 1,
+          termsTotal: terms.length,
+          currentUrl: term || "(all)",
+          jobPagesProcessed: out.length,
+          detail: `Careerjet: completed term ${i + 1}/${terms.length} (${collected} found)`,
+        });
+      }
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: out, error: message };
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/careerjet/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "careerjet-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "Careerjet public search API extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/careerjet/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/fourdayweek/README.md

@@ -0,0 +1,10 @@
+# fourdayweek-extractor
+
+Pulls listings from the public [4 Day Week API](https://4dayweek.io/api/jobs).
+
+- No authentication required.
+- Paginated JSON (up to 3 pages). Filters client-side by title + stack tags
+  against pipeline search terms.
+- No description in listings — links point to `https://4dayweek.io/job/{slug}`.
+- Rich metadata: schedule type, work-life score, salary, tech stack, level.
+- Caps results per term via the `fourdayweekMaxJobsPerTerm` setting (default 100).

extractors/fourdayweek/manifest.ts

@@ -0,0 +1,226 @@
+/**
+ * 4 Day Week public jobs API.
+ *
+ * https://4dayweek.io/api/jobs?page=N
+ *
+ * No auth. Paginated JSON. No description in listing response —
+ * we link to https://4dayweek.io/job/{slug} for details.
+ * Supports category filtering server-side; we also filter
+ * client-side by title + stack tags against pipeline search terms.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+const API_URL = "https://4dayweek.io/api/jobs";
+const MAX_PAGES = 3;
+
+interface FdwCompany {
+  name?: string;
+  slug?: string;
+  logo_url?: string;
+}
+
+interface FdwRemoteAllowed {
+  country?: string;
+  continent?: string;
+  is_primary?: boolean;
+}
+
+interface FdwStackItem {
+  name?: string;
+  slug?: string;
+}
+
+interface FdwJob {
+  id?: string;
+  title?: string;
+  slug?: string;
+  company_name?: string;
+  company?: FdwCompany;
+  work_arrangement?: string;
+  remote_allowed?: FdwRemoteAllowed[];
+  timezones?: string[];
+  posted?: number;
+  schedule_type?: string;
+  stack?: FdwStackItem[];
+  category?: string;
+  level?: string;
+  salary?: string;
+  salary_lower?: number;
+  salary_upper?: number;
+  salary_currency?: string;
+  salary_period?: string;
+  is_expired?: boolean;
+  work_life_score?: number;
+}
+
+interface FdwResponse {
+  jobs?: FdwJob[];
+  total?: number;
+  page?: number;
+  has_more?: boolean;
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed || undefined;
+}
+
+function matchesTerm(job: FdwJob, term: string): boolean {
+  const lower = term.toLowerCase();
+  if (job.title?.toLowerCase().includes(lower)) return true;
+  if (job.category?.toLowerCase().includes(lower)) return true;
+  if (
+    Array.isArray(job.stack) &&
+    job.stack.some(
+      (s) => typeof s.name === "string" && s.name.toLowerCase().includes(lower),
+    )
+  )
+    return true;
+  return false;
+}
+
+function formatSchedule(raw: string | undefined): string {
+  if (!raw) return "4-day week";
+  return raw.replace(/_/g, " ");
+}
+
+function formatLocation(job: FdwJob): string {
+  const countries = Array.isArray(job.remote_allowed)
+    ? job.remote_allowed
+        .map((r) => r.country)
+        .filter((c): c is string => typeof c === "string")
+    : [];
+  if (countries.length > 0) return countries.join(", ");
+  return job.work_arrangement === "remote" ? "Remote" : "Unknown";
+}
+
+function formatSalary(job: FdwJob): string | undefined {
+  if (job.salary) return job.salary;
+  if (job.salary_lower == null && job.salary_upper == null) return undefined;
+  const cur = job.salary_currency ?? "USD";
+  const period = job.salary_period ?? "year";
+  if (job.salary_lower != null && job.salary_upper != null) {
+    return `${cur} ${(job.salary_lower / 100).toLocaleString()}–${(job.salary_upper / 100).toLocaleString()} / ${period}`;
+  }
+  const val = job.salary_lower ?? job.salary_upper;
+  return val != null
+    ? `${cur} ${(val / 100).toLocaleString()} / ${period}`
+    : undefined;
+}
+
+function mapJob(raw: FdwJob): CreateJobInput | null {
+  const slug = asString(raw.slug);
+  if (!slug) return null;
+
+  const jobUrl = `https://4dayweek.io/job/${slug}`;
+  const stackTags = Array.isArray(raw.stack)
+    ? raw.stack
+        .map((s) => s.name)
+        .filter((n): n is string => typeof n === "string")
+    : [];
+
+  return {
+    source: "fourdayweek",
+    sourceJobId: raw.id ?? slug,
+    title: asString(raw.title) ?? "Unknown Title",
+    employer: raw.company?.name ?? raw.company_name ?? "Unknown Employer",
+    jobUrl,
+    applicationLink: jobUrl,
+    location: formatLocation(raw),
+    isRemote: raw.work_arrangement === "remote",
+    jobType: formatSchedule(raw.schedule_type),
+    companyLogo: raw.company?.logo_url ?? undefined,
+    datePosted:
+      typeof raw.posted === "number"
+        ? new Date(raw.posted * 1000).toISOString()
+        : undefined,
+    salary: formatSalary(raw),
+    disciplines: stackTags.length > 0 ? stackTags.join(", ") : undefined,
+    companyIndustry: asString(raw.category),
+  };
+}
+
+async function fetchPage(page: number): Promise<FdwResponse> {
+  const url = `${API_URL}?page=${page}`;
+  const response = await fetch(url, {
+    headers: { Accept: "application/json" },
+  });
+  if (!response.ok) {
+    throw new Error(`4 Day Week request failed with status ${response.status}`);
+  }
+  return (await response.json()) as FdwResponse;
+}
+
+export const manifest: ExtractorManifest = {
+  id: "fourdayweek",
+  displayName: "4 Day Week",
+  providesSources: ["fourdayweek"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const maxJobs = context.settings.fourdayweekMaxJobsPerTerm
+      ? Number.parseInt(context.settings.fourdayweekMaxJobsPerTerm, 10)
+      : 100;
+
+    const terms = context.searchTerms.length > 0 ? context.searchTerms : [];
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    try {
+      for (let page = 1; page <= MAX_PAGES; page += 1) {
+        if (context.shouldCancel?.()) break;
+        if (out.length >= maxJobs * Math.max(terms.length, 1)) break;
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: 0,
+          termsTotal: 1,
+          currentUrl: `page ${page}`,
+          detail: `4 Day Week: fetching page ${page}`,
+        });
+
+        const body = await fetchPage(page);
+        const jobs = Array.isArray(body.jobs) ? body.jobs : [];
+
+        if (jobs.length === 0) break;
+
+        for (const raw of jobs) {
+          if (raw.is_expired) continue;
+          if (terms.length > 0 && !terms.some((t) => matchesTerm(raw, t))) {
+            continue;
+          }
+          const mapped = mapJob(raw);
+          if (!mapped) continue;
+          const key = mapped.sourceJobId || mapped.jobUrl;
+          if (seen.has(key)) continue;
+          seen.add(key);
+          out.push(mapped);
+        }
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: 0,
+          termsTotal: 1,
+          currentUrl: `page ${page}`,
+          jobPagesProcessed: out.length,
+          detail: `4 Day Week: page ${page} done (${out.length} matched so far)`,
+        });
+
+        if (!body.has_more) break;
+      }
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: out, error: message };
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/fourdayweek/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "fourdayweek-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "4 Day Week public jobs API extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/fourdayweek/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/greenhouse/README.md

@@ -0,0 +1,8 @@
+# greenhouse-extractor
+
+Public Greenhouse Job Boards via
+`GET https://boards-api.greenhouse.io/v1/boards/{company}/jobs?content=true`.
+
+- No auth.
+- Configure target slugs via `greenhouseCompanies` (comma/newline) or
+  `GREENHOUSE_COMPANIES` env (e.g. `airbnb,stripe`).

extractors/greenhouse/manifest.ts

@@ -0,0 +1,188 @@
+/**
+ * Greenhouse public job boards API.
+ *
+ * https://developers.greenhouse.io/job-board.html
+ *   GET https://boards-api.greenhouse.io/v1/boards/{company}/jobs?content=true
+ *
+ * No auth. Each entry in `greenhouseCompanies` is fetched independently.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+interface GhDepartment {
+  id?: number;
+  name?: string;
+}
+interface GhMetadata {
+  name?: string;
+  value?: unknown;
+}
+interface GhJob {
+  id?: number;
+  title?: string;
+  absolute_url?: string;
+  internal_job_id?: number;
+  updated_at?: string;
+  requisition_id?: string | null;
+  location?: { name?: string };
+  content?: string; // HTML, may be entity-encoded
+  metadata?: GhMetadata[];
+  departments?: GhDepartment[];
+  offices?: Array<{ name?: string }>;
+}
+interface GhResponse {
+  jobs?: GhJob[];
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed ? trimmed : undefined;
+}
+
+function readCompanies(raw: string | undefined): string[] {
+  if (!raw) return [];
+  try {
+    const parsed = JSON.parse(raw);
+    if (Array.isArray(parsed)) {
+      return parsed
+        .map((entry) => (typeof entry === "string" ? entry.trim() : ""))
+        .filter(Boolean);
+    }
+  } catch {
+    // fall through
+  }
+  return raw
+    .split(/[\n,;|]+/)
+    .map((entry) => entry.trim())
+    .filter(Boolean);
+}
+
+function decodeHtmlEntities(value: string): string {
+  return value
+    .replace(/&amp;/g, "&")
+    .replace(/&lt;/g, "<")
+    .replace(/&gt;/g, ">")
+    .replace(/&quot;/g, '"')
+    .replace(/&#39;/g, "'")
+    .replace(/&nbsp;/g, " ");
+}
+
+function mapJob(job: GhJob, company: string): CreateJobInput | null {
+  const jobUrl = asString(job.absolute_url);
+  if (!jobUrl) return null;
+
+  const employer = company
+    .split(/[-_]/)
+    .filter(Boolean)
+    .map((part) => part.charAt(0).toUpperCase() + part.slice(1))
+    .join(" ");
+  const officeNames =
+    job.offices
+      ?.map((office) => asString(office.name))
+      .filter((name): name is string => Boolean(name)) ?? [];
+  const departmentNames =
+    job.departments
+      ?.map((dept) => asString(dept.name))
+      .filter((name): name is string => Boolean(name)) ?? [];
+
+  const description = job.content ? decodeHtmlEntities(job.content) : undefined;
+
+  return {
+    source: "greenhouse",
+    sourceJobId: job.id != null ? String(job.id) : undefined,
+    title: asString(job.title) ?? "Unknown Title",
+    employer: employer || company,
+    jobUrl,
+    applicationLink: jobUrl,
+    location:
+      asString(job.location?.name) ?? (officeNames.join("; ") || undefined),
+    jobFunction:
+      departmentNames.length > 0 ? departmentNames.join(", ") : undefined,
+    datePosted: asString(job.updated_at),
+    jobDescription: description,
+  };
+}
+
+async function fetchCompany(company: string): Promise<GhJob[]> {
+  const url = `https://boards-api.greenhouse.io/v1/boards/${encodeURIComponent(company)}/jobs?content=true`;
+  const response = await fetch(url, {
+    headers: { Accept: "application/json" },
+  });
+  if (response.status === 404) return [];
+  if (!response.ok) {
+    throw new Error(
+      `Greenhouse request for "${company}" failed with status ${response.status}`,
+    );
+  }
+  const body = (await response.json()) as GhResponse;
+  return Array.isArray(body.jobs) ? body.jobs : [];
+}
+
+export const manifest: ExtractorManifest = {
+  id: "greenhouse",
+  displayName: "Greenhouse (ATS)",
+  providesSources: ["greenhouse"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const companies = readCompanies(context.settings.greenhouseCompanies);
+    if (companies.length === 0) {
+      return {
+        success: true,
+        jobs: [],
+        error:
+          "No Greenhouse companies configured. Set GREENHOUSE_COMPANIES or the greenhouseCompanies setting (comma- or newline-separated slugs).",
+      };
+    }
+
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    try {
+      for (let i = 0; i < companies.length; i += 1) {
+        if (context.shouldCancel?.()) break;
+        const company = companies[i];
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i,
+          termsTotal: companies.length,
+          currentUrl: company,
+          detail: `Greenhouse: ${company} (${i + 1}/${companies.length})`,
+        });
+
+        let added = 0;
+        const jobs = await fetchCompany(company);
+        for (const job of jobs) {
+          const mapped = mapJob(job, company);
+          if (!mapped) continue;
+          const key = mapped.sourceJobId || mapped.jobUrl;
+          if (seen.has(key)) continue;
+          seen.add(key);
+          out.push(mapped);
+          added += 1;
+        }
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i + 1,
+          termsTotal: companies.length,
+          currentUrl: company,
+          jobPagesProcessed: out.length,
+          detail: `Greenhouse: ${company} → ${added} jobs (${out.length} total)`,
+        });
+      }
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: out, error: message };
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/greenhouse/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "greenhouse-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "Greenhouse public ATS extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/greenhouse/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/himalayas/README.md

@@ -0,0 +1,10 @@
+# himalayas-extractor
+
+Pulls listings from the public [Himalayas API](https://himalayas.app/jobs/api).
+
+- No authentication required.
+- Paginates with `limit` + `offset` (50 per page, up to 5 pages / 250 jobs).
+- No server-side search — filters client-side by matching title + categories
+  against each pipeline search term.
+- All listings are flagged `isRemote: true`.
+- Caps results per term via the `himalayasMaxJobsPerTerm` setting (default 100).

extractors/himalayas/manifest.ts

@@ -0,0 +1,195 @@
+/**
+ * Himalayas public remote-jobs API.
+ *
+ * https://himalayas.app/jobs/api?limit=N&offset=M
+ *
+ * No auth. Returns up to `limit` results per call. No server-side
+ * search — we paginate and filter client-side by title + categories.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+const API_URL = "https://himalayas.app/jobs/api";
+const PAGE_SIZE = 50;
+const MAX_PAGES = 5;
+
+interface HimalayasJob {
+  title?: string;
+  excerpt?: string;
+  companyName?: string;
+  companySlug?: string;
+  companyLogo?: string;
+  employmentType?: string;
+  minSalary?: number | null;
+  maxSalary?: number | null;
+  currency?: string;
+  seniority?: string[];
+  locationRestrictions?: string[];
+  timezoneRestrictions?: number[];
+  categories?: string[];
+  parentCategories?: string[];
+  description?: string;
+  pubDate?: number;
+  expiryDate?: number;
+  applicationLink?: string;
+  guid?: string;
+}
+
+interface HimalayasResponse {
+  jobs?: HimalayasJob[];
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed || undefined;
+}
+
+function matchesTerm(job: HimalayasJob, term: string): boolean {
+  const lower = term.toLowerCase();
+  if (job.title?.toLowerCase().includes(lower)) return true;
+  if (
+    Array.isArray(job.categories) &&
+    job.categories.some(
+      (c) =>
+        typeof c === "string" &&
+        c.toLowerCase().replace(/-/g, " ").includes(lower),
+    )
+  )
+    return true;
+  return false;
+}
+
+function formatSalary(job: HimalayasJob): string | undefined {
+  if (job.minSalary == null && job.maxSalary == null) return undefined;
+  const cur = job.currency ?? "USD";
+  if (job.minSalary != null && job.maxSalary != null) {
+    return `${cur} ${job.minSalary.toLocaleString()}–${job.maxSalary.toLocaleString()}`;
+  }
+  const val = job.minSalary ?? job.maxSalary;
+  return val != null ? `${cur} ${val.toLocaleString()}` : undefined;
+}
+
+function mapJob(raw: HimalayasJob): CreateJobInput | null {
+  const jobUrl = asString(raw.applicationLink) ?? asString(raw.guid);
+  if (!jobUrl) return null;
+
+  const categories = Array.isArray(raw.categories)
+    ? raw.categories.filter(
+        (c): c is string => typeof c === "string" && c.length > 0,
+      )
+    : [];
+
+  const locations = Array.isArray(raw.locationRestrictions)
+    ? raw.locationRestrictions.filter(
+        (l): l is string => typeof l === "string" && l.length > 0,
+      )
+    : [];
+
+  const datePosted =
+    typeof raw.pubDate === "number"
+      ? new Date(raw.pubDate * 1000).toISOString()
+      : undefined;
+
+  return {
+    source: "himalayas",
+    sourceJobId: asString(raw.guid),
+    title: asString(raw.title) ?? "Unknown Title",
+    employer: asString(raw.companyName) ?? "Unknown Employer",
+    jobUrl,
+    applicationLink: jobUrl,
+    location: locations.length > 0 ? locations.join(", ") : "Remote",
+    isRemote: true,
+    jobType: asString(raw.employmentType),
+    companyLogo: asString(raw.companyLogo),
+    datePosted,
+    salary: formatSalary(raw),
+    jobDescription: asString(raw.description),
+    disciplines: categories.length > 0 ? categories.join(", ") : undefined,
+  };
+}
+
+async function fetchPage(
+  offset: number,
+  limit: number,
+): Promise<HimalayasJob[]> {
+  const url = `${API_URL}?limit=${limit}&offset=${offset}`;
+  const response = await fetch(url, {
+    headers: { Accept: "application/json" },
+  });
+  if (!response.ok) {
+    throw new Error(`Himalayas request failed with status ${response.status}`);
+  }
+  const body = (await response.json()) as HimalayasResponse;
+  return Array.isArray(body.jobs) ? body.jobs : [];
+}
+
+export const manifest: ExtractorManifest = {
+  id: "himalayas",
+  displayName: "Himalayas",
+  providesSources: ["himalayas"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const maxJobs = context.settings.himalayasMaxJobsPerTerm
+      ? Number.parseInt(context.settings.himalayasMaxJobsPerTerm, 10)
+      : 100;
+
+    const terms = context.searchTerms.length > 0 ? context.searchTerms : [];
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    try {
+      for (let page = 0; page < MAX_PAGES; page += 1) {
+        if (context.shouldCancel?.()) break;
+        if (out.length >= maxJobs * Math.max(terms.length, 1)) break;
+
+        const offset = page * PAGE_SIZE;
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: 0,
+          termsTotal: 1,
+          currentUrl: `offset ${offset}`,
+          detail: `Himalayas: fetching page ${page + 1}`,
+        });
+
+        const raw = await fetchPage(offset, PAGE_SIZE);
+        if (raw.length === 0) break;
+
+        for (const item of raw) {
+          if (terms.length > 0 && !terms.some((t) => matchesTerm(item, t))) {
+            continue;
+          }
+          const mapped = mapJob(item);
+          if (!mapped) continue;
+          const key = mapped.sourceJobId || mapped.jobUrl;
+          if (seen.has(key)) continue;
+          seen.add(key);
+          out.push(mapped);
+        }
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: 0,
+          termsTotal: 1,
+          currentUrl: `offset ${offset}`,
+          jobPagesProcessed: out.length,
+          detail: `Himalayas: page ${page + 1} done (${out.length} matched so far)`,
+        });
+
+        if (raw.length < PAGE_SIZE) break;
+      }
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: out, error: message };
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/himalayas/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "himalayas-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "Himalayas public remote-jobs API extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/himalayas/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/jobicy/README.md

@@ -0,0 +1,8 @@
+# jobicy-extractor
+
+Pulls remote jobs from the public [Jobicy v2 feed](https://jobicy.com/api/v2/remote-jobs).
+
+- No authentication required.
+- Each pipeline `searchTerm` is sent as a `tag`; without terms we fetch the
+  generic remote feed.
+- Caps results via the `jobicyMaxJobsPerTerm` setting (default 100).

extractors/jobicy/manifest.ts

@@ -0,0 +1,186 @@
+/**
+ * Jobicy remote-jobs feed.
+ *
+ * Public, unauthenticated JSON endpoint:
+ *   https://jobicy.com/api/v2/remote-jobs?count=50
+ *
+ * The feed is intentionally remote-only; we still pass each `searchTerm` as a
+ * `tag` so the same pipeline-level term iteration drives results. We do *not*
+ * try to invent a country filter — Jobicy postings are remote-friendly by
+ * design and the registry already restricts ukOnly extractors elsewhere.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+const API_URL = "https://jobicy.com/api/v2/remote-jobs";
+
+interface JobicyRawJob {
+  id?: number | string;
+  url?: string;
+  jobTitle?: string;
+  companyName?: string;
+  companyLogo?: string;
+  jobIndustry?: string[] | string;
+  jobType?: string[] | string;
+  jobGeo?: string;
+  jobLevel?: string;
+  jobExcerpt?: string;
+  jobDescription?: string;
+  pubDate?: string;
+  annualSalaryMin?: number | string;
+  annualSalaryMax?: number | string;
+  salaryCurrency?: string;
+}
+
+interface JobicyResponse {
+  jobs?: JobicyRawJob[];
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed ? trimmed : undefined;
+}
+
+function joinList(value: unknown): string | undefined {
+  if (Array.isArray(value)) {
+    const cleaned = value
+      .map((item) => (typeof item === "string" ? item.trim() : ""))
+      .filter(Boolean);
+    return cleaned.length > 0 ? cleaned.join(", ") : undefined;
+  }
+  return asString(value);
+}
+
+function toNumberOrUndefined(value: unknown): number | undefined {
+  if (typeof value === "number" && Number.isFinite(value)) return value;
+  if (typeof value === "string") {
+    const parsed = Number.parseFloat(value);
+    return Number.isFinite(parsed) ? parsed : undefined;
+  }
+  return undefined;
+}
+
+function mapJob(raw: JobicyRawJob): CreateJobInput | null {
+  const jobUrl = asString(raw.url);
+  if (!jobUrl) return null;
+
+  const employer = asString(raw.companyName) ?? "Unknown Employer";
+  const title = asString(raw.jobTitle) ?? "Unknown Title";
+  const minSalary = toNumberOrUndefined(raw.annualSalaryMin);
+  const maxSalary = toNumberOrUndefined(raw.annualSalaryMax);
+
+  return {
+    source: "jobicy",
+    sourceJobId: raw.id != null ? String(raw.id) : undefined,
+    title,
+    employer,
+    jobUrl,
+    applicationLink: jobUrl,
+    location: asString(raw.jobGeo) ?? "Remote",
+    isRemote: true,
+    jobType: joinList(raw.jobType),
+    jobLevel: asString(raw.jobLevel),
+    companyIndustry: joinList(raw.jobIndustry),
+    companyLogo: asString(raw.companyLogo),
+    datePosted: asString(raw.pubDate),
+    jobDescription: asString(raw.jobDescription) ?? asString(raw.jobExcerpt),
+    salaryMinAmount: minSalary,
+    salaryMaxAmount: maxSalary,
+    salaryCurrency: asString(raw.salaryCurrency),
+    salaryInterval: minSalary || maxSalary ? "yearly" : undefined,
+  };
+}
+
+async function fetchJobicy(
+  tag: string | null,
+  count: number,
+): Promise<JobicyRawJob[]> {
+  const url = new URL(API_URL);
+  url.searchParams.set("count", String(Math.min(Math.max(count, 1), 50)));
+  if (tag) url.searchParams.set("tag", tag);
+
+  const response = await fetch(url.toString(), {
+    headers: { Accept: "application/json" },
+  });
+  if (!response.ok) {
+    throw new Error(`Jobicy request failed with status ${response.status}`);
+  }
+  const body = (await response.json()) as JobicyResponse;
+  return Array.isArray(body.jobs) ? body.jobs : [];
+}
+
+export const manifest: ExtractorManifest = {
+  id: "jobicy",
+  displayName: "Jobicy (Remote)",
+  providesSources: ["jobicy"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const maxJobsPerTerm = context.settings.jobicyMaxJobsPerTerm
+      ? Number.parseInt(context.settings.jobicyMaxJobsPerTerm, 10)
+      : 100;
+
+    const terms = context.searchTerms.length > 0 ? context.searchTerms : [null];
+
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    try {
+      for (let i = 0; i < terms.length; i += 1) {
+        if (context.shouldCancel?.()) break;
+        const term = terms[i];
+        const tag = term ? term.trim().toLowerCase() : null;
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i,
+          termsTotal: terms.length,
+          currentUrl: tag ?? "(all remote)",
+          detail: `Jobicy: term ${i + 1}/${terms.length}`,
+        });
+
+        // Jobicy caps `count` at 50 per call; loop until we either hit the
+        // requested cap or the feed runs out (length < take).
+        let collected = 0;
+        let safetyHops = 0;
+        while (collected < maxJobsPerTerm && safetyHops < 10) {
+          const take = Math.min(50, maxJobsPerTerm - collected);
+          const raw = await fetchJobicy(tag, take);
+          if (raw.length === 0) break;
+          for (const item of raw) {
+            const mapped = mapJob(item);
+            if (!mapped) continue;
+            const key = mapped.sourceJobId || mapped.jobUrl;
+            if (seen.has(key)) continue;
+            seen.add(key);
+            out.push(mapped);
+            collected += 1;
+            if (collected >= maxJobsPerTerm) break;
+          }
+          if (raw.length < take) break;
+          safetyHops += 1;
+        }
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i + 1,
+          termsTotal: terms.length,
+          currentUrl: tag ?? "(all remote)",
+          jobPagesProcessed: out.length,
+          detail: `Jobicy: completed term ${i + 1}/${terms.length} (${collected} found)`,
+        });
+      }
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: out, error: message };
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/jobicy/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "jobicy-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "Jobicy remote-jobs feed extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/jobicy/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/jooble/README.md

@@ -0,0 +1,7 @@
+# jooble-extractor
+
+[Jooble](https://jooble.org/api/about) aggregator API extractor.
+
+- Requires `JOOBLE_API_KEY` (`joobleApiKey` setting).
+- Iterates `searchTerms`; uses the first `searchCities` token as `location`.
+- Capped per term via `joobleMaxJobsPerTerm` (default 100).

extractors/jooble/manifest.ts

@@ -0,0 +1,177 @@
+/**
+ * Jooble aggregator API.
+ *
+ * https://jooble.org/api/about — `POST https://jooble.org/api/{key}` with a
+ * JSON body of `{ keywords, location, page, ResultOnPage }`.
+ *
+ * Requires JOOBLE_API_KEY (`joobleApiKey` setting).
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+const API_BASE = "https://jooble.org/api";
+
+interface JoobleJob {
+  id?: number | string;
+  title?: string;
+  location?: string;
+  snippet?: string;
+  salary?: string;
+  source?: string;
+  type?: string;
+  link?: string;
+  company?: string;
+  updated?: string;
+}
+interface JoobleResponse {
+  totalCount?: number;
+  jobs?: JoobleJob[];
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed ? trimmed : undefined;
+}
+
+function mapJob(raw: JoobleJob): CreateJobInput | null {
+  const jobUrl = asString(raw.link);
+  if (!jobUrl) return null;
+  return {
+    source: "jooble",
+    sourceJobId: raw.id != null ? String(raw.id) : undefined,
+    title: asString(raw.title) ?? "Unknown Title",
+    employer: asString(raw.company) ?? "Unknown Employer",
+    jobUrl,
+    applicationLink: jobUrl,
+    location: asString(raw.location),
+    jobType: asString(raw.type),
+    salary: asString(raw.salary),
+    datePosted: asString(raw.updated),
+    jobDescription: asString(raw.snippet),
+    companyDescription: asString(raw.source),
+  };
+}
+
+async function fetchPage(args: {
+  apiKey: string;
+  keywords: string;
+  location?: string;
+  page: number;
+  resultOnPage: number;
+}): Promise<JoobleResponse> {
+  const response = await fetch(
+    `${API_BASE}/${encodeURIComponent(args.apiKey)}`,
+    {
+      method: "POST",
+      headers: {
+        "Content-Type": "application/json",
+        Accept: "application/json",
+      },
+      body: JSON.stringify({
+        keywords: args.keywords,
+        location: args.location ?? "",
+        page: String(args.page),
+        ResultOnPage: String(args.resultOnPage),
+      }),
+    },
+  );
+  if (!response.ok) {
+    throw new Error(`Jooble request failed with status ${response.status}`);
+  }
+  return (await response.json()) as JoobleResponse;
+}
+
+export const manifest: ExtractorManifest = {
+  id: "jooble",
+  displayName: "Jooble",
+  providesSources: ["jooble"],
+  requiredEnvVars: ["JOOBLE_API_KEY"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const apiKey =
+      context.settings.joobleApiKey?.trim() ||
+      process.env.JOOBLE_API_KEY?.trim();
+    if (!apiKey) {
+      return {
+        success: false,
+        jobs: [],
+        error: "Jooble extractor requires JOOBLE_API_KEY",
+      };
+    }
+
+    const maxJobsPerTerm = context.settings.joobleMaxJobsPerTerm
+      ? Number.parseInt(context.settings.joobleMaxJobsPerTerm, 10)
+      : 100;
+    const resultOnPage = 50;
+
+    const terms = context.searchTerms.length > 0 ? context.searchTerms : [""];
+    const location =
+      context.settings.searchCities?.split("|")[0]?.trim() || undefined;
+
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    try {
+      for (let i = 0; i < terms.length; i += 1) {
+        if (context.shouldCancel?.()) break;
+        const term = terms[i].trim();
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i,
+          termsTotal: terms.length,
+          currentUrl: term || "(all)",
+          detail: `Jooble: term ${i + 1}/${terms.length}`,
+        });
+
+        let collected = 0;
+        let page = 1;
+        while (collected < maxJobsPerTerm && page < 50) {
+          if (context.shouldCancel?.()) break;
+          const body = await fetchPage({
+            apiKey,
+            keywords: term,
+            location,
+            page,
+            resultOnPage,
+          });
+          const items = Array.isArray(body.jobs) ? body.jobs : [];
+          if (items.length === 0) break;
+          for (const raw of items) {
+            const mapped = mapJob(raw);
+            if (!mapped) continue;
+            const key = mapped.sourceJobId || mapped.jobUrl;
+            if (seen.has(key)) continue;
+            seen.add(key);
+            out.push(mapped);
+            collected += 1;
+            if (collected >= maxJobsPerTerm) break;
+          }
+          if (items.length < resultOnPage) break;
+          page += 1;
+        }
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i + 1,
+          termsTotal: terms.length,
+          currentUrl: term || "(all)",
+          jobPagesProcessed: out.length,
+          detail: `Jooble: completed term ${i + 1}/${terms.length} (${collected} found)`,
+        });
+      }
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: out, error: message };
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/jooble/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "jooble-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "Jooble aggregator API extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/jooble/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/lever/README.md

@@ -0,0 +1,9 @@
+# lever-extractor
+
+Public Lever ATS feeds via `GET https://api.lever.co/v0/postings/{company}?mode=json`.
+
+- No auth.
+- Configure target slugs through `leverCompanies` (comma/newline list) or
+  `LEVER_COMPANIES` env (e.g. `figma,plaid,ramp`).
+- Pulls every active posting from each company; pipeline filters handle terms
+  and country gating.

extractors/lever/manifest.ts

@@ -0,0 +1,182 @@
+/**
+ * Lever public job postings API.
+ *
+ * https://github.com/lever/postings-api/blob/master/README.md
+ *   GET https://api.lever.co/v0/postings/{company}?mode=json
+ *
+ * No auth. We iterate `leverCompanies` (set in Settings or LEVER_COMPANIES env)
+ * and pull every active posting; downstream filtering by `searchTerms` /
+ * country happens in the pipeline.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+interface LeverCategories {
+  team?: string;
+  department?: string;
+  commitment?: string;
+  location?: string;
+  allLocations?: string[];
+}
+interface LeverPosting {
+  id?: string;
+  text?: string;
+  hostedUrl?: string;
+  applyUrl?: string;
+  description?: string;
+  descriptionPlain?: string;
+  categories?: LeverCategories;
+  createdAt?: number;
+  workplaceType?: string;
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed ? trimmed : undefined;
+}
+
+function readCompanies(raw: string | undefined): string[] {
+  if (!raw) return [];
+  try {
+    const parsed = JSON.parse(raw);
+    if (Array.isArray(parsed)) {
+      return parsed
+        .map((entry) =>
+          typeof entry === "string" ? entry.trim().toLowerCase() : "",
+        )
+        .filter((entry) => entry.length > 0);
+    }
+  } catch {
+    // fall through to delimited-list parsing below
+  }
+  return raw
+    .split(/[\n,;|]+/)
+    .map((entry) => entry.trim().toLowerCase())
+    .filter(Boolean);
+}
+
+function locationFor(posting: LeverPosting): string | undefined {
+  const cats = posting.categories;
+  if (!cats) return undefined;
+  if (Array.isArray(cats.allLocations) && cats.allLocations.length > 0) {
+    return cats.allLocations.filter(Boolean).join("; ");
+  }
+  return asString(cats.location);
+}
+
+function mapPosting(
+  posting: LeverPosting,
+  company: string,
+): CreateJobInput | null {
+  const jobUrl = asString(posting.hostedUrl);
+  if (!jobUrl) return null;
+  const employer = company
+    .split("-")
+    .filter(Boolean)
+    .map((part) => part.charAt(0).toUpperCase() + part.slice(1))
+    .join(" ");
+  return {
+    source: "lever",
+    sourceJobId: asString(posting.id),
+    title: asString(posting.text) ?? "Unknown Title",
+    employer: employer || company,
+    jobUrl,
+    applicationLink: asString(posting.applyUrl) ?? jobUrl,
+    location: locationFor(posting),
+    jobType: asString(posting.categories?.commitment),
+    jobFunction: asString(posting.categories?.team),
+    companyIndustry: asString(posting.categories?.department),
+    isRemote:
+      posting.workplaceType?.toLowerCase() === "remote" ? true : undefined,
+    datePosted:
+      typeof posting.createdAt === "number"
+        ? new Date(posting.createdAt).toISOString()
+        : undefined,
+    jobDescription:
+      asString(posting.descriptionPlain) ?? asString(posting.description),
+  };
+}
+
+async function fetchCompany(company: string): Promise<LeverPosting[]> {
+  const url = `https://api.lever.co/v0/postings/${encodeURIComponent(company)}?mode=json`;
+  const response = await fetch(url, {
+    headers: { Accept: "application/json" },
+  });
+  if (response.status === 404) return [];
+  if (!response.ok) {
+    throw new Error(
+      `Lever request for "${company}" failed with status ${response.status}`,
+    );
+  }
+  const body = (await response.json()) as unknown;
+  return Array.isArray(body) ? (body as LeverPosting[]) : [];
+}
+
+export const manifest: ExtractorManifest = {
+  id: "lever",
+  displayName: "Lever (ATS)",
+  providesSources: ["lever"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const companies = readCompanies(context.settings.leverCompanies);
+    if (companies.length === 0) {
+      return {
+        success: true,
+        jobs: [],
+        error:
+          "No Lever companies configured. Set LEVER_COMPANIES or the leverCompanies setting (comma- or newline-separated slugs).",
+      };
+    }
+
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    try {
+      for (let i = 0; i < companies.length; i += 1) {
+        if (context.shouldCancel?.()) break;
+        const company = companies[i];
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i,
+          termsTotal: companies.length,
+          currentUrl: company,
+          detail: `Lever: ${company} (${i + 1}/${companies.length})`,
+        });
+
+        let added = 0;
+        const postings = await fetchCompany(company);
+        for (const posting of postings) {
+          const mapped = mapPosting(posting, company);
+          if (!mapped) continue;
+          const key = mapped.sourceJobId || mapped.jobUrl;
+          if (seen.has(key)) continue;
+          seen.add(key);
+          out.push(mapped);
+          added += 1;
+        }
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i + 1,
+          termsTotal: companies.length,
+          currentUrl: company,
+          jobPagesProcessed: out.length,
+          detail: `Lever: ${company} → ${added} jobs (${out.length} total)`,
+        });
+      }
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: out, error: message };
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/lever/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "lever-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "Lever public ATS extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/lever/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/reed/README.md

@@ -0,0 +1,8 @@
+# reed-extractor
+
+[Reed.co.uk Jobseeker API](https://www.reed.co.uk/developers/jobseeker).
+
+- Requires `REED_API_KEY` (`reedApiKey` setting), used as the HTTP Basic
+  username.
+- UK-only: gated via `isSourceAllowedForCountry`.
+- Capped per term via `reedMaxJobsPerTerm` (default 100).

extractors/reed/manifest.ts

@@ -0,0 +1,188 @@
+/**
+ * Reed.co.uk Jobseeker API.
+ *
+ * https://www.reed.co.uk/developers/jobseeker
+ *   GET https://www.reed.co.uk/api/1.0/search?...
+ *   HTTP Basic with the API key as the username and an empty password.
+ *
+ * Requires REED_API_KEY (`reedApiKey` setting). The catalog gates this source
+ * to UK only via `isSourceAllowedForCountry`.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+const API_URL = "https://www.reed.co.uk/api/1.0/search";
+
+interface ReedJob {
+  jobId?: number;
+  jobTitle?: string;
+  employerName?: string;
+  employerProfileUrl?: string;
+  jobDescription?: string;
+  jobUrl?: string;
+  locationName?: string;
+  date?: string;
+  expirationDate?: string;
+  applications?: number;
+  currency?: string;
+  minimumSalary?: number;
+  maximumSalary?: number;
+  yearlyMinimumSalary?: number;
+  yearlyMaximumSalary?: number;
+}
+interface ReedResponse {
+  totalResults?: number;
+  results?: ReedJob[];
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed ? trimmed : undefined;
+}
+
+function mapJob(raw: ReedJob): CreateJobInput | null {
+  const jobUrl = asString(raw.jobUrl);
+  if (!jobUrl) return null;
+  return {
+    source: "reed",
+    sourceJobId: raw.jobId != null ? String(raw.jobId) : undefined,
+    title: asString(raw.jobTitle) ?? "Unknown Title",
+    employer: asString(raw.employerName) ?? "Unknown Employer",
+    employerUrl: asString(raw.employerProfileUrl),
+    jobUrl,
+    applicationLink: jobUrl,
+    location: asString(raw.locationName),
+    datePosted: asString(raw.date),
+    deadline: asString(raw.expirationDate),
+    jobDescription: asString(raw.jobDescription),
+    salaryMinAmount:
+      typeof raw.minimumSalary === "number" ? raw.minimumSalary : undefined,
+    salaryMaxAmount:
+      typeof raw.maximumSalary === "number" ? raw.maximumSalary : undefined,
+    salaryCurrency: asString(raw.currency) ?? "GBP",
+    salaryInterval: raw.yearlyMinimumSalary != null ? "yearly" : undefined,
+  };
+}
+
+async function fetchPage(args: {
+  apiKey: string;
+  keywords: string;
+  locationName?: string;
+  resultsToTake: number;
+  resultsToSkip: number;
+}): Promise<ReedResponse> {
+  const url = new URL(API_URL);
+  url.searchParams.set("keywords", args.keywords);
+  if (args.locationName)
+    url.searchParams.set("locationName", args.locationName);
+  url.searchParams.set("resultsToTake", String(args.resultsToTake));
+  url.searchParams.set("resultsToSkip", String(args.resultsToSkip));
+
+  const auth = Buffer.from(`${args.apiKey}:`).toString("base64");
+  const response = await fetch(url.toString(), {
+    headers: {
+      Accept: "application/json",
+      Authorization: `Basic ${auth}`,
+    },
+  });
+  if (!response.ok) {
+    throw new Error(`Reed request failed with status ${response.status}`);
+  }
+  return (await response.json()) as ReedResponse;
+}
+
+export const manifest: ExtractorManifest = {
+  id: "reed",
+  displayName: "Reed",
+  providesSources: ["reed"],
+  requiredEnvVars: ["REED_API_KEY"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const apiKey =
+      context.settings.reedApiKey?.trim() || process.env.REED_API_KEY?.trim();
+    if (!apiKey) {
+      return {
+        success: false,
+        jobs: [],
+        error: "Reed extractor requires REED_API_KEY",
+      };
+    }
+
+    const maxJobsPerTerm = context.settings.reedMaxJobsPerTerm
+      ? Number.parseInt(context.settings.reedMaxJobsPerTerm, 10)
+      : 100;
+    // Reed accepts up to 100 per page.
+    const pageSize = Math.min(100, maxJobsPerTerm);
+
+    const terms = context.searchTerms.length > 0 ? context.searchTerms : [""];
+    const locationName =
+      context.settings.searchCities?.split("|")[0]?.trim() || undefined;
+
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    try {
+      for (let i = 0; i < terms.length; i += 1) {
+        if (context.shouldCancel?.()) break;
+        const term = terms[i].trim();
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i,
+          termsTotal: terms.length,
+          currentUrl: term || "(all)",
+          detail: `Reed: term ${i + 1}/${terms.length}`,
+        });
+
+        let collected = 0;
+        let resultsToSkip = 0;
+        while (collected < maxJobsPerTerm) {
+          if (context.shouldCancel?.()) break;
+          const body = await fetchPage({
+            apiKey,
+            keywords: term,
+            locationName,
+            resultsToTake: pageSize,
+            resultsToSkip,
+          });
+          const items = Array.isArray(body.results) ? body.results : [];
+          if (items.length === 0) break;
+          for (const raw of items) {
+            const mapped = mapJob(raw);
+            if (!mapped) continue;
+            const key = mapped.sourceJobId || mapped.jobUrl;
+            if (seen.has(key)) continue;
+            seen.add(key);
+            out.push(mapped);
+            collected += 1;
+            if (collected >= maxJobsPerTerm) break;
+          }
+          if (items.length < pageSize) break;
+          resultsToSkip += pageSize;
+          if (resultsToSkip > 5000) break;
+        }
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i + 1,
+          termsTotal: terms.length,
+          currentUrl: term || "(all)",
+          jobPagesProcessed: out.length,
+          detail: `Reed: completed term ${i + 1}/${terms.length} (${collected} found)`,
+        });
+      }
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: out, error: message };
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/reed/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "reed-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "Reed.co.uk Jobseeker API extractor (UK only)",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/reed/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/remoteok/README.md

@@ -0,0 +1,15 @@
+# remoteok-extractor
+
+Pulls listings from the public [Remote OK feed](https://remoteok.com/api).
+
+- No authentication required.
+- The endpoint returns the entire active board in a single JSON array; the
+  first element is metadata/legal text. We fetch once and apply each pipeline
+  `searchTerm` as a case-insensitive filter over the job's `position` and
+  `tags`.
+- Caps results per term via the `remoteokMaxJobsPerTerm` setting (default 100).
+- Listings are flagged `isRemote: true` and labelled "Remote" if Remote OK
+  doesn't supply a city.
+- Per Remote OK's TOS we send a descriptive User-Agent and preserve the legal
+  notice on the response. If you publish results, please link back to the
+  original posting URLs.

extractors/remoteok/manifest.ts

@@ -0,0 +1,190 @@
+/**
+ * Remote OK public feed.
+ *
+ * https://remoteok.com/api — single JSON endpoint that returns the entire
+ * active remote-jobs board in one shot. The first array element is metadata
+ * (legal/attribution); jobs follow.
+ *
+ * No auth, no server-side pagination, no per-term query — we fetch once per
+ * pipeline run and apply each `searchTerm` as a case-insensitive filter over
+ * `position` + `tags` so the orchestrator's per-term iteration still works.
+ *
+ * Per Remote OK's TOS we send a descriptive User-Agent so they can identify
+ * traffic; we do not strip the legal/attribution element from the response.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+const API_URL = "https://remoteok.com/api";
+const USER_AGENT =
+  "Mozilla/5.0 (compatible; JobOps/1.0; +https://github.com/) job-search pipeline";
+
+interface RemoteOkJob {
+  id?: string | number;
+  slug?: string;
+  position?: string;
+  company?: string;
+  company_logo?: string;
+  logo?: string;
+  location?: string;
+  tags?: string[];
+  description?: string;
+  url?: string;
+  apply_url?: string;
+  date?: string;
+  epoch?: number;
+  salary_min?: number;
+  salary_max?: number;
+}
+
+interface RemoteOkLegalEntry {
+  legal?: string;
+  last_updated?: number;
+}
+
+type RemoteOkResponseEntry = RemoteOkJob | RemoteOkLegalEntry;
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed ? trimmed : undefined;
+}
+
+function isJobEntry(entry: RemoteOkResponseEntry): entry is RemoteOkJob {
+  return (
+    "id" in entry || "position" in entry || "url" in entry || "slug" in entry
+  );
+}
+
+function tagMatchesTerm(job: RemoteOkJob, normalizedTerm: string): boolean {
+  if (!normalizedTerm) return true;
+  const haystack = [
+    job.position ?? "",
+    ...(Array.isArray(job.tags) ? job.tags : []),
+  ]
+    .join(" ")
+    .toLowerCase();
+  return haystack.includes(normalizedTerm);
+}
+
+function mapJob(job: RemoteOkJob): CreateJobInput | null {
+  const jobUrl = asString(job.url) ?? asString(job.apply_url);
+  if (!jobUrl) return null;
+
+  // Remote OK reports salary as raw numbers; 0 means "not specified".
+  const minSalary =
+    typeof job.salary_min === "number" && job.salary_min > 0
+      ? job.salary_min
+      : undefined;
+  const maxSalary =
+    typeof job.salary_max === "number" && job.salary_max > 0
+      ? job.salary_max
+      : undefined;
+
+  const tags = Array.isArray(job.tags)
+    ? job.tags.filter((tag): tag is string => typeof tag === "string")
+    : [];
+
+  return {
+    source: "remoteok",
+    sourceJobId: job.id != null ? String(job.id) : asString(job.slug),
+    title: asString(job.position) ?? "Unknown Title",
+    employer: asString(job.company) ?? "Unknown Employer",
+    jobUrl,
+    applicationLink: asString(job.apply_url) ?? jobUrl,
+    location: asString(job.location) ?? "Remote",
+    isRemote: true,
+    datePosted: asString(job.date),
+    jobDescription: asString(job.description),
+    companyLogo: asString(job.company_logo) ?? asString(job.logo),
+    disciplines: tags.length > 0 ? tags.join(", ") : undefined,
+    salaryMinAmount: minSalary,
+    salaryMaxAmount: maxSalary,
+    salaryCurrency: minSalary || maxSalary ? "USD" : undefined,
+    salaryInterval: minSalary || maxSalary ? "yearly" : undefined,
+  };
+}
+
+async function fetchAll(): Promise<RemoteOkJob[]> {
+  const response = await fetch(API_URL, {
+    headers: {
+      Accept: "application/json",
+      "User-Agent": USER_AGENT,
+    },
+  });
+  if (!response.ok) {
+    throw new Error(`Remote OK request failed with status ${response.status}`);
+  }
+  const body = (await response.json()) as RemoteOkResponseEntry[];
+  if (!Array.isArray(body)) return [];
+  return body.filter(isJobEntry);
+}
+
+export const manifest: ExtractorManifest = {
+  id: "remoteok",
+  displayName: "Remote OK",
+  providesSources: ["remoteok"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const maxJobsPerTerm = context.settings.remoteokMaxJobsPerTerm
+      ? Number.parseInt(context.settings.remoteokMaxJobsPerTerm, 10)
+      : 100;
+
+    const terms = context.searchTerms.length > 0 ? context.searchTerms : [""];
+
+    let allJobs: RemoteOkJob[];
+    try {
+      allJobs = await fetchAll();
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: [], error: message };
+    }
+
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    for (let i = 0; i < terms.length; i += 1) {
+      if (context.shouldCancel?.()) break;
+      const term = terms[i].trim();
+      const normalizedTerm = term.toLowerCase();
+      context.onProgress?.({
+        phase: "list",
+        termsProcessed: i,
+        termsTotal: terms.length,
+        currentUrl: term || "(all remote)",
+        detail: `Remote OK: term ${i + 1}/${terms.length}`,
+      });
+
+      let collected = 0;
+      for (const job of allJobs) {
+        if (collected >= maxJobsPerTerm) break;
+        if (!tagMatchesTerm(job, normalizedTerm)) continue;
+        const mapped = mapJob(job);
+        if (!mapped) continue;
+        const key = mapped.sourceJobId || mapped.jobUrl;
+        if (seen.has(key)) continue;
+        seen.add(key);
+        out.push(mapped);
+        collected += 1;
+      }
+
+      context.onProgress?.({
+        phase: "list",
+        termsProcessed: i + 1,
+        termsTotal: terms.length,
+        currentUrl: term || "(all remote)",
+        jobPagesProcessed: out.length,
+        detail: `Remote OK: completed term ${i + 1}/${terms.length} (${collected} matched)`,
+      });
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/remoteok/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "remoteok-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "Remote OK public job-feed extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/remoteok/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/remotive/README.md

@@ -0,0 +1,9 @@
+# remotive-extractor
+
+Pulls listings from the public [Remotive API](https://remotive.com/api/remote-jobs).
+
+- No authentication required.
+- Each pipeline `searchTerm` is passed as the `search` query parameter;
+  without terms we fetch the generic remote feed.
+- Caps results per term via the `remotiveMaxJobsPerTerm` setting (default 100).
+- All listings are flagged `isRemote: true`.

extractors/remotive/manifest.ts

@@ -0,0 +1,153 @@
+/**
+ * Remotive public remote-jobs API.
+ *
+ * https://remotive.com/api/remote-jobs?limit=N&search=term
+ *
+ * No auth. Returns up to `limit` results per call with a `search` keyword
+ * filter. We iterate pipeline search terms as the `search` parameter.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+const API_URL = "https://remotive.com/api/remote-jobs";
+
+interface RemotiveJob {
+  id?: number;
+  url?: string;
+  title?: string;
+  company_name?: string;
+  company_logo?: string;
+  category?: string;
+  tags?: string[];
+  job_type?: string;
+  publication_date?: string;
+  candidate_required_location?: string;
+  salary?: string;
+  description?: string;
+}
+
+interface RemotiveResponse {
+  jobs?: RemotiveJob[];
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed ? trimmed : undefined;
+}
+
+function normalizeJobType(raw: string | undefined): string | undefined {
+  if (!raw) return undefined;
+  return raw.replace(/_/g, " ").trim() || undefined;
+}
+
+function mapJob(raw: RemotiveJob): CreateJobInput | null {
+  const jobUrl = asString(raw.url);
+  if (!jobUrl) return null;
+
+  const tags = Array.isArray(raw.tags)
+    ? raw.tags.filter((t): t is string => typeof t === "string" && t.length > 0)
+    : [];
+
+  return {
+    source: "remotive",
+    sourceJobId: raw.id != null ? String(raw.id) : undefined,
+    title: asString(raw.title) ?? "Unknown Title",
+    employer: asString(raw.company_name) ?? "Unknown Employer",
+    jobUrl,
+    applicationLink: jobUrl,
+    location: asString(raw.candidate_required_location) ?? "Remote",
+    isRemote: true,
+    jobType: normalizeJobType(raw.job_type),
+    companyIndustry: asString(raw.category),
+    companyLogo: asString(raw.company_logo),
+    datePosted: asString(raw.publication_date),
+    salary: asString(raw.salary),
+    jobDescription: asString(raw.description),
+    disciplines: tags.length > 0 ? tags.join(", ") : undefined,
+  };
+}
+
+async function fetchJobs(
+  search: string | null,
+  limit: number,
+): Promise<RemotiveJob[]> {
+  const url = new URL(API_URL);
+  url.searchParams.set("limit", String(Math.min(Math.max(limit, 1), 100)));
+  if (search) url.searchParams.set("search", search);
+
+  const response = await fetch(url.toString(), {
+    headers: { Accept: "application/json" },
+  });
+  if (!response.ok) {
+    throw new Error(`Remotive request failed with status ${response.status}`);
+  }
+  const body = (await response.json()) as RemotiveResponse;
+  return Array.isArray(body.jobs) ? body.jobs : [];
+}
+
+export const manifest: ExtractorManifest = {
+  id: "remotive",
+  displayName: "Remotive",
+  providesSources: ["remotive"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const maxJobsPerTerm = context.settings.remotiveMaxJobsPerTerm
+      ? Number.parseInt(context.settings.remotiveMaxJobsPerTerm, 10)
+      : 100;
+
+    const terms = context.searchTerms.length > 0 ? context.searchTerms : [null];
+
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    try {
+      for (let i = 0; i < terms.length; i += 1) {
+        if (context.shouldCancel?.()) break;
+        const term = terms[i];
+        const search = term ? term.trim() : null;
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i,
+          termsTotal: terms.length,
+          currentUrl: search ?? "(all remote)",
+          detail: `Remotive: term ${i + 1}/${terms.length}`,
+        });
+
+        const raw = await fetchJobs(search, maxJobsPerTerm);
+        let collected = 0;
+        for (const item of raw) {
+          if (collected >= maxJobsPerTerm) break;
+          const mapped = mapJob(item);
+          if (!mapped) continue;
+          const key = mapped.sourceJobId || mapped.jobUrl;
+          if (seen.has(key)) continue;
+          seen.add(key);
+          out.push(mapped);
+          collected += 1;
+        }
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i + 1,
+          termsTotal: terms.length,
+          currentUrl: search ?? "(all remote)",
+          jobPagesProcessed: out.length,
+          detail: `Remotive: completed term ${i + 1}/${terms.length} (${collected} found)`,
+        });
+      }
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: out, error: message };
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/remotive/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "remotive-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "Remotive public remote-jobs API extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/remotive/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/themuse/README.md

@@ -0,0 +1,8 @@
+# themuse-extractor
+
+Pulls postings from [The Muse public jobs API](https://www.themuse.com/developers/api/v2).
+
+- Works without auth, but `themuseApiKey` (`THEMUSE_API_KEY`) raises rate limits.
+- Each pipeline `searchTerm` is sent as a `category`. The Muse's first
+  `searchCities` token is forwarded as `location`.
+- Capped per term via `themuseMaxJobsPerTerm` (default 100).

extractors/themuse/manifest.ts

@@ -0,0 +1,224 @@
+/**
+ * The Muse public jobs API.
+ *
+ * https://www.themuse.com/api/public/jobs?page=0&category=...&location=...
+ *
+ * The endpoint works without auth but is heavily rate-limited; an API key
+ * (THEMUSE_API_KEY / `themuseApiKey` setting) lifts that. We pass each pipeline
+ * search term as a `category` to keep parity with how other extractors iterate
+ * search terms; if your role doesn't map to a Muse category it'll still match
+ * because Muse falls back to generic listings.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+const API_URL = "https://www.themuse.com/api/public/jobs";
+
+interface MuseLocation {
+  name?: string;
+}
+interface MuseCompany {
+  name?: string;
+  short_name?: string;
+}
+interface MuseRefs {
+  landing_page?: string;
+}
+interface MuseJob {
+  id?: number;
+  name?: string;
+  publication_date?: string;
+  type?: string;
+  contents?: string;
+  short_description?: string;
+  locations?: MuseLocation[];
+  company?: MuseCompany;
+  refs?: MuseRefs;
+}
+interface MuseResponse {
+  page?: number;
+  page_count?: number;
+  results?: MuseJob[];
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed ? trimmed : undefined;
+}
+
+function joinLocations(
+  locations: MuseLocation[] | undefined,
+): string | undefined {
+  if (!locations || locations.length === 0) return undefined;
+  const cleaned = locations
+    .map((entry) => asString(entry.name))
+    .filter((name): name is string => Boolean(name));
+  return cleaned.length > 0 ? cleaned.join("; ") : undefined;
+}
+
+function isRemoteFromLocations(
+  locations: MuseLocation[] | undefined,
+): boolean | undefined {
+  if (!locations || locations.length === 0) return undefined;
+  return locations.some((loc) =>
+    typeof loc.name === "string"
+      ? /\bflexible|remote\b/i.test(loc.name)
+      : false,
+  );
+}
+
+// The Muse `category` filter expects an exact, Title-Cased category name (e.g.
+// "Software Engineer", "Engineering"). User-supplied search terms are commonly
+// lowercase free-text, which the API silently ignores and returns zero results.
+// Title-case the term so common values map to real categories; if the term
+// still doesn't match a category the extractor will fall back to no filter.
+function toMuseCategory(term: string): string | undefined {
+  const trimmed = term.trim();
+  if (!trimmed) return undefined;
+  return trimmed
+    .toLowerCase()
+    .split(/\s+/)
+    .map((part) => part.charAt(0).toUpperCase() + part.slice(1))
+    .join(" ");
+}
+
+function mapJob(raw: MuseJob): CreateJobInput | null {
+  const jobUrl = asString(raw.refs?.landing_page);
+  if (!jobUrl) return null;
+
+  return {
+    source: "themuse",
+    sourceJobId: raw.id != null ? String(raw.id) : undefined,
+    title: asString(raw.name) ?? "Unknown Title",
+    employer: asString(raw.company?.name) ?? "Unknown Employer",
+    jobUrl,
+    applicationLink: jobUrl,
+    location: joinLocations(raw.locations),
+    isRemote: isRemoteFromLocations(raw.locations),
+    jobType: asString(raw.type),
+    datePosted: asString(raw.publication_date),
+    jobDescription:
+      asString(raw.contents) ?? asString(raw.short_description) ?? undefined,
+  };
+}
+
+async function fetchPage(args: {
+  apiKey?: string;
+  page: number;
+  category?: string;
+  location?: string;
+}): Promise<MuseResponse> {
+  const url = new URL(API_URL);
+  url.searchParams.set("page", String(args.page));
+  if (args.category) url.searchParams.set("category", args.category);
+  if (args.location) url.searchParams.set("location", args.location);
+  if (args.apiKey) url.searchParams.set("api_key", args.apiKey);
+
+  const response = await fetch(url.toString(), {
+    headers: { Accept: "application/json" },
+  });
+  if (!response.ok) {
+    throw new Error(`The Muse request failed with status ${response.status}`);
+  }
+  return (await response.json()) as MuseResponse;
+}
+
+export const manifest: ExtractorManifest = {
+  id: "themuse",
+  displayName: "The Muse",
+  providesSources: ["themuse"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const apiKey = context.settings.themuseApiKey?.trim() || undefined;
+    const maxJobsPerTerm = context.settings.themuseMaxJobsPerTerm
+      ? Number.parseInt(context.settings.themuseMaxJobsPerTerm, 10)
+      : 100;
+
+    const terms = context.searchTerms.length > 0 ? context.searchTerms : [""];
+    const locationHint =
+      context.settings.searchCities?.split("|")[0]?.trim() || undefined;
+
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    try {
+      for (let i = 0; i < terms.length; i += 1) {
+        if (context.shouldCancel?.()) break;
+        const term = terms[i].trim();
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i,
+          termsTotal: terms.length,
+          currentUrl: term || "(all)",
+          detail: `The Muse: term ${i + 1}/${terms.length}`,
+        });
+
+        let collected = 0;
+        let page = 0;
+        let pageCount = Number.POSITIVE_INFINITY;
+        // The Muse returns pageCount; cap pages defensively to avoid runaway
+        // loops if the API misbehaves. We try the term as a category first and,
+        // if the very first page is empty, drop the category filter once so an
+        // unknown category doesn't silently nuke the entire term.
+        let categoryToUse: string | undefined = toMuseCategory(term);
+        let droppedCategory = false;
+        while (collected < maxJobsPerTerm && page < pageCount && page < 100) {
+          if (context.shouldCancel?.()) break;
+          const body = await fetchPage({
+            apiKey,
+            page,
+            category: categoryToUse,
+            location: locationHint,
+          });
+          if (typeof body.page_count === "number") {
+            pageCount = body.page_count;
+          }
+          const results = Array.isArray(body.results) ? body.results : [];
+          if (results.length === 0) {
+            if (page === 0 && categoryToUse && !droppedCategory) {
+              categoryToUse = undefined;
+              droppedCategory = true;
+              pageCount = Number.POSITIVE_INFINITY;
+              continue;
+            }
+            break;
+          }
+
+          for (const item of results) {
+            const mapped = mapJob(item);
+            if (!mapped) continue;
+            const key = mapped.sourceJobId || mapped.jobUrl;
+            if (seen.has(key)) continue;
+            seen.add(key);
+            out.push(mapped);
+            collected += 1;
+            if (collected >= maxJobsPerTerm) break;
+          }
+          page += 1;
+        }
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i + 1,
+          termsTotal: terms.length,
+          currentUrl: term || "(all)",
+          jobPagesProcessed: out.length,
+          detail: `The Muse: completed term ${i + 1}/${terms.length} (${collected} found)`,
+        });
+      }
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: out, error: message };
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/themuse/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "themuse-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "The Muse public jobs API extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/themuse/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/usajobs/README.md

@@ -0,0 +1,22 @@
+# usajobs-extractor
+
+US-government job listings via the public
+[USAJOBS Search API](https://developer.usajobs.gov/api-reference/get-api-search) (`GET /api/Search` on `data.usajobs.gov`).
+
+## Getting API access
+
+1. Open the [USAJOBS Developer Site](https://developer.usajobs.gov/) and complete **USAJOBS API Access Request** (sign in / register as required).
+2. After approval, USAJOBS emails your **API key** to the address you used—use that value for `USAJOBS_API_KEY`.
+3. Set `USAJOBS_USER_AGENT` to a **real contact email** (same one you used for registration is typical). This is required by their terms, not optional branding—the HTTP `User-Agent` header must identify you.
+
+Reference material on the developer site includes API Reference, tutorials, and code lists (locations, pay plans, etc.).
+
+## Configuration
+
+| Env / setting | Required | Notes |
+| --- | --- | --- |
+| `USAJOBS_API_KEY` / `usajobsApiKey` | yes | From the email USAJOBS sends after API access is granted |
+| `USAJOBS_USER_AGENT` / `usajobsUserAgent` | yes | Real contact email per USAJOBS TOS |
+| `usajobsMaxJobsPerTerm` | no | Per-term cap (default 100) |
+
+The orchestrator's country gating restricts this source to the United States.

extractors/usajobs/manifest.ts

@@ -0,0 +1,263 @@
+/**
+ * USAJOBS public search API.
+ *
+ * https://developer.usajobs.gov/api-reference/get-api-search
+ *
+ * Requires:
+ *   - USAJOBS_API_KEY (`usajobsApiKey` setting)
+ *   - USAJOBS_USER_AGENT — must be a real contact email per their TOS
+ *
+ * The orchestrator already gates this source to United States via
+ * `isSourceAllowedForCountry`, so we don't re-validate country here.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+const API_URL = "https://data.usajobs.gov/api/Search";
+
+interface UsaJobsLocation {
+  LocationName?: string;
+  CountryCode?: string;
+}
+interface UsaJobsRemuneration {
+  MinimumRange?: string;
+  MaximumRange?: string;
+  RateIntervalCode?: string;
+}
+interface UsaJobsDescriptor {
+  PositionID?: string;
+  PositionTitle?: string;
+  PositionURI?: string;
+  ApplyURI?: string[];
+  PositionLocationDisplay?: string;
+  PositionLocation?: UsaJobsLocation[];
+  OrganizationName?: string;
+  DepartmentName?: string;
+  PublicationStartDate?: string;
+  PositionStartDate?: string;
+  PositionEndDate?: string;
+  PositionRemuneration?: UsaJobsRemuneration[];
+  UserArea?: { Details?: { JobSummary?: string } };
+  PositionSchedule?: Array<{ Name?: string }>;
+}
+interface UsaJobsSearchResultItem {
+  MatchedObjectDescriptor?: UsaJobsDescriptor;
+}
+interface UsaJobsSearchResult {
+  SearchResult?: {
+    SearchResultCountAll?: number;
+    SearchResultItems?: UsaJobsSearchResultItem[];
+  };
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed ? trimmed : undefined;
+}
+
+function toNumberOrUndefined(value: unknown): number | undefined {
+  if (typeof value === "number" && Number.isFinite(value)) return value;
+  if (typeof value === "string") {
+    const parsed = Number.parseFloat(value);
+    return Number.isFinite(parsed) ? parsed : undefined;
+  }
+  return undefined;
+}
+
+function mapInterval(code: string | undefined): string | undefined {
+  if (!code) return undefined;
+  switch (code.toLowerCase()) {
+    case "py":
+    case "pa":
+      return "yearly";
+    case "ph":
+      return "hourly";
+    case "pd":
+      return "daily";
+    case "pm":
+      return "monthly";
+    case "pw":
+      return "weekly";
+    default:
+      return undefined;
+  }
+}
+
+function mapJob(item: UsaJobsSearchResultItem): CreateJobInput | null {
+  const descriptor = item.MatchedObjectDescriptor;
+  if (!descriptor) return null;
+  const jobUrl = asString(descriptor.PositionURI);
+  if (!jobUrl) return null;
+
+  const remuneration = descriptor.PositionRemuneration?.[0];
+  const min = toNumberOrUndefined(remuneration?.MinimumRange);
+  const max = toNumberOrUndefined(remuneration?.MaximumRange);
+  const interval = mapInterval(remuneration?.RateIntervalCode);
+  const applyArr = descriptor.ApplyURI;
+  const applicationLink =
+    Array.isArray(applyArr) && applyArr.length > 0
+      ? (asString(applyArr[0]) ?? jobUrl)
+      : jobUrl;
+
+  return {
+    source: "usajobs",
+    sourceJobId: asString(descriptor.PositionID),
+    title: asString(descriptor.PositionTitle) ?? "Unknown Title",
+    employer:
+      asString(descriptor.OrganizationName) ??
+      asString(descriptor.DepartmentName) ??
+      "U.S. Federal Government",
+    jobUrl,
+    applicationLink,
+    location: asString(descriptor.PositionLocationDisplay),
+    datePosted: asString(descriptor.PublicationStartDate),
+    deadline: asString(descriptor.PositionEndDate),
+    jobDescription: asString(descriptor.UserArea?.Details?.JobSummary),
+    jobType: descriptor.PositionSchedule?.[0]?.Name?.trim() || undefined,
+    salaryMinAmount: min,
+    salaryMaxAmount: max,
+    salaryCurrency: min || max ? "USD" : undefined,
+    salaryInterval: interval,
+  };
+}
+
+async function fetchPage(args: {
+  apiKey: string;
+  userAgent: string;
+  keyword: string;
+  locationName?: string;
+  page: number;
+  resultsPerPage: number;
+}): Promise<UsaJobsSearchResult> {
+  const url = new URL(API_URL);
+  url.searchParams.set("Keyword", args.keyword);
+  if (args.locationName) {
+    url.searchParams.set("LocationName", args.locationName);
+  }
+  url.searchParams.set("ResultsPerPage", String(args.resultsPerPage));
+  url.searchParams.set("Page", String(args.page));
+  url.searchParams.set("SortField", "OpenDate");
+  url.searchParams.set("SortDirection", "Desc");
+
+  const response = await fetch(url.toString(), {
+    headers: {
+      Host: "data.usajobs.gov",
+      "User-Agent": args.userAgent,
+      "Authorization-Key": args.apiKey,
+      Accept: "application/json",
+    },
+  });
+  if (!response.ok) {
+    throw new Error(`USAJOBS request failed with status ${response.status}`);
+  }
+  return (await response.json()) as UsaJobsSearchResult;
+}
+
+export const manifest: ExtractorManifest = {
+  id: "usajobs",
+  displayName: "USAJOBS",
+  providesSources: ["usajobs"],
+  requiredEnvVars: ["USAJOBS_API_KEY", "USAJOBS_USER_AGENT"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const apiKey =
+      context.settings.usajobsApiKey?.trim() ||
+      process.env.USAJOBS_API_KEY?.trim();
+    const userAgent =
+      context.settings.usajobsUserAgent?.trim() ||
+      process.env.USAJOBS_USER_AGENT?.trim();
+    if (!apiKey || !userAgent) {
+      return {
+        success: false,
+        jobs: [],
+        error:
+          "USAJOBS extractor requires USAJOBS_API_KEY and USAJOBS_USER_AGENT (a contact email)",
+      };
+    }
+
+    const maxJobsPerTerm = context.settings.usajobsMaxJobsPerTerm
+      ? Number.parseInt(context.settings.usajobsMaxJobsPerTerm, 10)
+      : 100;
+    // USAJOBS caps page size at 500, but smaller pages are friendlier on retry.
+    const resultsPerPage = 50;
+
+    const terms = context.searchTerms.length > 0 ? context.searchTerms : [""];
+    const locationName =
+      context.settings.searchCities?.split("|")[0]?.trim() || undefined;
+
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+
+    try {
+      for (let i = 0; i < terms.length; i += 1) {
+        if (context.shouldCancel?.()) break;
+        const term = terms[i].trim();
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i,
+          termsTotal: terms.length,
+          currentUrl: term || "(all)",
+          detail: `USAJOBS: term ${i + 1}/${terms.length}`,
+        });
+
+        let collected = 0;
+        let page = 1;
+        let total = Number.POSITIVE_INFINITY;
+        while (
+          collected < maxJobsPerTerm &&
+          (page - 1) * resultsPerPage < total &&
+          page < 200
+        ) {
+          if (context.shouldCancel?.()) break;
+          const body = await fetchPage({
+            apiKey,
+            userAgent,
+            keyword: term,
+            locationName,
+            page,
+            resultsPerPage,
+          });
+          if (typeof body.SearchResult?.SearchResultCountAll === "number") {
+            total = body.SearchResult.SearchResultCountAll;
+          }
+          const items = body.SearchResult?.SearchResultItems ?? [];
+          if (items.length === 0) break;
+          for (const item of items) {
+            const mapped = mapJob(item);
+            if (!mapped) continue;
+            const key = mapped.sourceJobId || mapped.jobUrl;
+            if (seen.has(key)) continue;
+            seen.add(key);
+            out.push(mapped);
+            collected += 1;
+            if (collected >= maxJobsPerTerm) break;
+          }
+          if (items.length < resultsPerPage) break;
+          page += 1;
+        }
+
+        context.onProgress?.({
+          phase: "list",
+          termsProcessed: i + 1,
+          termsTotal: terms.length,
+          currentUrl: term || "(all)",
+          jobPagesProcessed: out.length,
+          detail: `USAJOBS: completed term ${i + 1}/${terms.length} (${collected} found)`,
+        });
+      }
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: out, error: message };
+    }
+
+    return { success: true, jobs: out };
+  },
+};
+
+export default manifest;

extractors/usajobs/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "usajobs-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "USAJOBS public search API extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/usajobs/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/weworkremotely/README.md

@@ -0,0 +1,10 @@
+# weworkremotely-extractor
+
+Pulls listings from the public [We Work Remotely RSS feed](https://weworkremotely.com/remote-jobs.rss).
+
+- No authentication required.
+- Single RSS fetch returns all recent listings; we filter client-side
+  by matching title + skills + category against pipeline search terms.
+- Uses lightweight regex-based XML parsing (no external XML library).
+- All listings are flagged `isRemote: true`.
+- Caps results per term via the `weworkremotelyMaxJobsPerTerm` setting (default 100).

extractors/weworkremotely/manifest.ts

@@ -0,0 +1,192 @@
+/**
+ * We Work Remotely — public RSS feed.
+ *
+ * https://weworkremotely.com/remote-jobs.rss
+ *
+ * No auth. Returns all recent listings in a single XML feed.
+ * We filter client-side by matching title + skills + category
+ * against each pipeline search term.
+ *
+ * Title format from WWR: "Company Name: Job Title"
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+const RSS_URL = "https://weworkremotely.com/remote-jobs.rss";
+
+interface WwrItem {
+  title?: string;
+  link?: string;
+  guid?: string;
+  description?: string;
+  pubDate?: string;
+  region?: string;
+  country?: string;
+  skills?: string;
+  category?: string;
+  type?: string;
+  logoUrl?: string;
+}
+
+function xmlText(xml: string, tag: string): string | undefined {
+  const pattern = new RegExp(`<${tag}[^>]*>([\\s\\S]*?)</${tag}>`);
+  const match = xml.match(pattern);
+  if (!match?.[1]) return undefined;
+  return (
+    match[1].replace(/<!\[CDATA\[([\s\S]*?)\]\]>/g, "$1").trim() || undefined
+  );
+}
+
+function parseItems(xml: string): WwrItem[] {
+  const items: WwrItem[] = [];
+  const blocks = xml.match(/<item>([\s\S]*?)<\/item>/g) ?? [];
+
+  for (const raw of blocks) {
+    const block = raw.replace(/^<item>/, "").replace(/<\/item>$/, "");
+    const logoMatch = block.match(/media:content\s+url="([^"]+)"/);
+
+    items.push({
+      title: xmlText(block, "title"),
+      link: xmlText(block, "link"),
+      guid: xmlText(block, "guid"),
+      description: xmlText(block, "description"),
+      pubDate: xmlText(block, "pubDate"),
+      region: xmlText(block, "region"),
+      country: xmlText(block, "country"),
+      skills: xmlText(block, "skills"),
+      category: xmlText(block, "category"),
+      type: xmlText(block, "type"),
+      logoUrl: logoMatch?.[1],
+    });
+  }
+
+  return items;
+}
+
+function parseTitle(raw: string): { employer: string; title: string } {
+  const colonIdx = raw.indexOf(": ");
+  if (colonIdx > 0) {
+    return {
+      employer: raw.slice(0, colonIdx).trim(),
+      title: raw.slice(colonIdx + 2).trim(),
+    };
+  }
+  return { employer: "Unknown Employer", title: raw.trim() };
+}
+
+function matchesTerm(item: WwrItem, term: string): boolean {
+  const lower = term.toLowerCase();
+  if (item.title?.toLowerCase().includes(lower)) return true;
+  if (item.skills?.toLowerCase().includes(lower)) return true;
+  if (item.category?.toLowerCase().includes(lower)) return true;
+  return false;
+}
+
+function decodeHtmlEntities(html: string): string {
+  return html
+    .replace(/&amp;/g, "&")
+    .replace(/&lt;/g, "<")
+    .replace(/&gt;/g, ">");
+}
+
+function mapJob(item: WwrItem): CreateJobInput | null {
+  const jobUrl = item.link || item.guid;
+  if (!jobUrl) return null;
+
+  const rawTitle = item.title
+    ? decodeHtmlEntities(item.title)
+    : "Unknown Title";
+  const { employer, title } = parseTitle(rawTitle);
+
+  const location =
+    [item.region, item.country].filter(Boolean).join(" — ") || "Remote";
+
+  return {
+    source: "weworkremotely",
+    sourceJobId: item.guid ?? item.link,
+    title,
+    employer,
+    jobUrl,
+    applicationLink: jobUrl,
+    location,
+    isRemote: true,
+    jobType: item.type || undefined,
+    companyLogo: item.logoUrl,
+    datePosted: item.pubDate,
+    jobDescription: item.description
+      ? decodeHtmlEntities(item.description)
+      : undefined,
+    disciplines: item.skills || undefined,
+    companyIndustry: item.category || undefined,
+  };
+}
+
+export const manifest: ExtractorManifest = {
+  id: "weworkremotely",
+  displayName: "We Work Remotely",
+  providesSources: ["weworkremotely"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const maxJobs = context.settings.weworkremotelyMaxJobsPerTerm
+      ? Number.parseInt(context.settings.weworkremotelyMaxJobsPerTerm, 10)
+      : 100;
+
+    const terms = context.searchTerms.length > 0 ? context.searchTerms : [];
+
+    context.onProgress?.({
+      phase: "list",
+      termsProcessed: 0,
+      termsTotal: 1,
+      currentUrl: RSS_URL,
+      detail: "We Work Remotely: fetching RSS feed",
+    });
+
+    try {
+      const response = await fetch(RSS_URL, {
+        headers: { Accept: "application/rss+xml, application/xml, text/xml" },
+      });
+      if (!response.ok) {
+        throw new Error(`WWR RSS failed with status ${response.status}`);
+      }
+      const xml = await response.text();
+      const items = parseItems(xml);
+
+      const seen = new Set<string>();
+      const out: CreateJobInput[] = [];
+
+      for (const item of items) {
+        if (out.length >= maxJobs * Math.max(terms.length, 1)) break;
+        if (terms.length > 0 && !terms.some((t) => matchesTerm(item, t))) {
+          continue;
+        }
+        const mapped = mapJob(item);
+        if (!mapped) continue;
+        const key = mapped.sourceJobId || mapped.jobUrl;
+        if (seen.has(key)) continue;
+        seen.add(key);
+        out.push(mapped);
+      }
+
+      context.onProgress?.({
+        phase: "list",
+        termsProcessed: 1,
+        termsTotal: 1,
+        currentUrl: RSS_URL,
+        jobPagesProcessed: out.length,
+        detail: `We Work Remotely: ${out.length} matched from ${items.length} total`,
+      });
+
+      return { success: true, jobs: out };
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Unknown error";
+      return { success: false, jobs: [], error: message };
+    }
+  },
+};
+
+export default manifest;

extractors/weworkremotely/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "weworkremotely-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "We Work Remotely RSS feed extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/weworkremotely/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

extractors/workday/README.md

@@ -0,0 +1,25 @@
+# workday-extractor
+
+Public Workday career sites via the JSON CXS endpoint
+`POST {tenantUrl}/wday/cxs/{tenant}/{site}/jobs`.
+
+## Configuration
+
+Set `workdayTenants` (or `WORKDAY_TENANTS` env). Each entry is either:
+
+1. A career-site URL we'll auto-parse, e.g.
+   `https://nvidia.wd5.myworkdayjobs.com/en-US/NVIDIAExternalCareerSite`
+2. A JSON object with explicit fields:
+
+```json
+{
+  "company": "NVIDIA",
+  "tenantUrl": "https://nvidia.wd5.myworkdayjobs.com",
+  "tenant": "nvidia",
+  "site": "NVIDIAExternalCareerSite",
+  "locale": "en-US"
+}
+```
+
+Multiple entries are separated by newlines or commas. Pipeline `searchTerms`
+are passed as the request `searchText`.

extractors/workday/manifest.ts

@@ -0,0 +1,263 @@
+/**
+ * Workday public career-site extractor.
+ *
+ * Workday tenants expose their public job board over a JSON CXS endpoint:
+ *   POST {tenantUrl}/wday/cxs/{tenant}/{site}/jobs
+ *   { appliedFacets: {}, limit: 20, offset: 0, searchText: "..." }
+ *
+ * `workdayTenants` accepts entries shaped as JSON objects (preferred) or as
+ * career-page URLs we parse on a best-effort basis. When we can't recover the
+ * tenant + site we skip the entry and continue.
+ */
+
+import type {
+  ExtractorManifest,
+  ExtractorRunResult,
+} from "@shared/types/extractors";
+import type { CreateJobInput } from "@shared/types/jobs";
+
+interface WorkdayTarget {
+  company: string;
+  tenantUrl: string;
+  tenant: string;
+  site: string;
+  locale?: string;
+}
+
+interface WorkdayJobPosting {
+  title?: string;
+  externalPath?: string;
+  locationsText?: string;
+  postedOn?: string;
+  bulletFields?: string[];
+}
+interface WorkdayResponse {
+  total?: number;
+  jobPostings?: WorkdayJobPosting[];
+}
+
+function asString(value: unknown): string | undefined {
+  if (typeof value !== "string") return undefined;
+  const trimmed = value.trim();
+  return trimmed ? trimmed : undefined;
+}
+
+function inferTenantFromHost(host: string): string | null {
+  // host looks like `acme.wd5.myworkdayjobs.com` → tenant "acme"
+  const match = host.match(/^([^.]+)\.wd\d+\.myworkdayjobs\.com$/i);
+  return match ? match[1] : null;
+}
+
+function parseTargetEntry(entry: string): WorkdayTarget | null {
+  const trimmed = entry.trim();
+  if (!trimmed) return null;
+  // First, try JSON.
+  try {
+    const parsed = JSON.parse(trimmed) as Partial<WorkdayTarget>;
+    if (
+      parsed &&
+      typeof parsed.company === "string" &&
+      typeof parsed.tenantUrl === "string" &&
+      typeof parsed.tenant === "string" &&
+      typeof parsed.site === "string"
+    ) {
+      return {
+        company: parsed.company,
+        tenantUrl: parsed.tenantUrl.replace(/\/$/, ""),
+        tenant: parsed.tenant,
+        site: parsed.site,
+        locale: typeof parsed.locale === "string" ? parsed.locale : undefined,
+      };
+    }
+  } catch {
+    // Fall through to URL parsing.
+  }
+
+  // URL form, e.g.
+  //   https://nvidia.wd5.myworkdayjobs.com/en-US/NVIDIAExternalCareerSite
+  try {
+    const url = new URL(trimmed);
+    const tenant = inferTenantFromHost(url.host);
+    if (!tenant) return null;
+    const segments = url.pathname.split("/").filter(Boolean);
+    if (segments.length < 2) return null;
+    const [maybeLocale, site] = segments;
+    return {
+      company: tenant,
+      tenantUrl: `${url.protocol}//${url.host}`,
+      tenant,
+      site,
+      locale: maybeLocale,
+    };
+  } catch {
+    return null;
+  }
+}
+
+function readTargets(raw: string | undefined): WorkdayTarget[] {
+  if (!raw) return [];
+  const out: WorkdayTarget[] = [];
+  // settings store stringifies JSON arrays; if we got a JSON array of strings
+  // we still need to parse each entry individually.
+  let entries: string[] = [];
+  try {
+    const parsed = JSON.parse(raw);
+    if (Array.isArray(parsed)) {
+      entries = parsed
+        .map((entry) =>
+          typeof entry === "string" ? entry : JSON.stringify(entry),
+        )
+        .filter(Boolean);
+    }
+  } catch {
+    entries = raw
+      .split(/\n+/)
+      .map((line) => line.trim())
+      .filter(Boolean);
+  }
+  if (entries.length === 0) {
+    entries = raw
+      .split(/\n+/)
+      .map((line) => line.trim())
+      .filter(Boolean);
+  }
+  for (const entry of entries) {
+    const target = parseTargetEntry(entry);
+    if (target) out.push(target);
+  }
+  return out;
+}
+
+function mapPosting(
+  posting: WorkdayJobPosting,
+  target: WorkdayTarget,
+): CreateJobInput | null {
+  const externalPath = asString(posting.externalPath);
+  if (!externalPath) return null;
+  const locale = target.locale ?? "en-US";
+  const jobUrl = `${target.tenantUrl}/${locale}/${target.site}${externalPath}`;
+  return {
+    source: "workday",
+    sourceJobId: externalPath,
+    title: asString(posting.title) ?? "Unknown Title",
+    employer: target.company,
+    jobUrl,
+    applicationLink: jobUrl,
+    location: asString(posting.locationsText),
+    datePosted: asString(posting.postedOn),
+    jobType: posting.bulletFields?.find((field) => field?.length)?.trim(),
+  };
+}
+
+async function fetchPage(args: {
+  target: WorkdayTarget;
+  searchText: string;
+  offset: number;
+  limit: number;
+}): Promise<WorkdayResponse> {
+  const url = `${args.target.tenantUrl}/wday/cxs/${encodeURIComponent(args.target.tenant)}/${encodeURIComponent(args.target.site)}/jobs`;
+  const response = await fetch(url, {
+    method: "POST",
+    headers: {
+      "Content-Type": "application/json",
+      Accept: "application/json",
+    },
+    body: JSON.stringify({
+      appliedFacets: {},
+      limit: args.limit,
+      offset: args.offset,
+      searchText: args.searchText,
+    }),
+  });
+  if (!response.ok) {
+    throw new Error(
+      `Workday request for "${args.target.company}" failed with status ${response.status}`,
+    );
+  }
+  return (await response.json()) as WorkdayResponse;
+}
+
+export const manifest: ExtractorManifest = {
+  id: "workday",
+  displayName: "Workday (ATS)",
+  providesSources: ["workday"],
+  async run(context): Promise<ExtractorRunResult> {
+    if (context.shouldCancel?.()) return { success: true, jobs: [] };
+
+    const targets = readTargets(context.settings.workdayTenants);
+    if (targets.length === 0) {
+      return {
+        success: true,
+        jobs: [],
+        error:
+          "No Workday tenants configured. Set WORKDAY_TENANTS or the workdayTenants setting to a list of career-site URLs (or JSON entries with company/tenantUrl/tenant/site).",
+      };
+    }
+
+    const terms = context.searchTerms.length > 0 ? context.searchTerms : [""];
+    const seen = new Set<string>();
+    const out: CreateJobInput[] = [];
+    const limit = 20;
+    const errors: string[] = [];
+
+    for (let t = 0; t < targets.length; t += 1) {
+      if (context.shouldCancel?.()) break;
+      const target = targets[t];
+      try {
+        for (let i = 0; i < terms.length; i += 1) {
+          if (context.shouldCancel?.()) break;
+          const term = terms[i].trim();
+          context.onProgress?.({
+            phase: "list",
+            termsProcessed: t * terms.length + i,
+            termsTotal: targets.length * terms.length,
+            currentUrl: `${target.company} (${term || "all"})`,
+            detail: `Workday: ${target.company} term ${i + 1}/${terms.length}`,
+          });
+
+          let offset = 0;
+          let total = Number.POSITIVE_INFINITY;
+          while (offset < total && offset < 1000) {
+            if (context.shouldCancel?.()) break;
+            const body = await fetchPage({
+              target,
+              searchText: term,
+              offset,
+              limit,
+            });
+            if (typeof body.total === "number") total = body.total;
+            const postings = Array.isArray(body.jobPostings)
+              ? body.jobPostings
+              : [];
+            if (postings.length === 0) break;
+            for (const posting of postings) {
+              const mapped = mapPosting(posting, target);
+              if (!mapped) continue;
+              const key = mapped.sourceJobId || mapped.jobUrl;
+              if (seen.has(key)) continue;
+              seen.add(key);
+              out.push(mapped);
+            }
+            offset += postings.length;
+            if (postings.length < limit) break;
+          }
+        }
+      } catch (error) {
+        const message =
+          error instanceof Error ? error.message : "Unknown error";
+        errors.push(`${target.company}: ${message}`);
+      }
+    }
+
+    if (out.length === 0 && errors.length > 0) {
+      return { success: false, jobs: out, error: errors.join("; ") };
+    }
+    return {
+      success: true,
+      jobs: out,
+      error: errors.length > 0 ? errors.join("; ") : undefined,
+    };
+  },
+};
+
+export default manifest;

extractors/workday/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "workday-extractor",
+  "version": "0.0.1",
+  "type": "module",
+  "description": "Workday public career-site extractor",
+  "main": "manifest.ts",
+  "dependencies": {
+    "job-ops-shared": "^1.0.0"
+  },
+  "devDependencies": {
+    "@types/node": "^24.0.0",
+    "typescript": "~5.9.0"
+  },
+  "scripts": {
+    "check:types": "tsc --noEmit"
+  }
+}

extractors/workday/tsconfig.json

@@ -0,0 +1,17 @@
+{
+  "compilerOptions": {
+    "module": "ESNext",
+    "moduleResolution": "bundler",
+    "target": "ES2022",
+    "outDir": "dist",
+    "strict": true,
+    "noUnusedLocals": false,
+    "lib": ["ES2022", "DOM"],
+    "types": ["node"],
+    "baseUrl": ".",
+    "paths": {
+      "@shared/*": ["../../shared/src/*"]
+    }
+  },
+  "include": ["./manifest.ts", "./src/**/*"]
+}

orchestrator/package.json

@@ -1,6 +1,6 @@
 {
   "name": "job-ops-orchestrator",
-  "version": "0.2.0",
+  "version": "0.2.1",
   "type": "module",
   "description": "Unified orchestrator for job application pipeline",
   "main": "src/server/index.ts",

orchestrator/src/client/api/client.ts

@@ -502,11 +502,33 @@ async function fetchApi<T>(
   options?: RequestInit,
 ): Promise<T> {
   const method = (options?.method || "GET").toUpperCase();
-  let authHeader = cachedBasicAuthCredentials
+  const activeCreds = getActiveBasicAuthCredentials();
-    ? encodeBasicAuthHeaderValue(cachedBasicAuthCredentials)
+  let authHeader = activeCreds ? encodeBasicAuthHeaderValue(activeCreds) : undefined;
-    : undefined;
   let authAttempt = 0;
-  let usernameHint = cachedBasicAuthCredentials?.username;
+  let usernameHint = activeCreds?.username;
+
+  const shouldPromptForAuth = (args: {
+    method: string;
+    endpoint: string;
+    response: Response;
+    parsed: ApiResponse<unknown> | LegacyApiResponse<unknown>;
+  }): boolean => {
+    if (!basicAuthPromptHandler) return false;
+    if (authAttempt >= 2) return false;
+    if (!isUnauthorizedResponse(args.response, args.parsed)) return false;
+
+    // By default we only prompt for write methods. However, some parts of the UI
+    // (Settings/Profile) require auth even for reads. Without prompting, those
+    // screens appear to "forget" saved data after refresh.
+    if (isWriteMethod(args.method)) return true;
+
+    const readAuthEndpoints = new Set<string>([
+      "/settings",
+      "/profile",
+      "/profiles",
+    ]);
+    return readAuthEndpoints.has(args.endpoint);
+  };
 
   while (true) {
     const { response, parsed } = await fetchAndParse(
@@ -515,12 +537,7 @@ async function fetchApi<T>(
       authHeader,
     );
 
-    if (
+    if (shouldPromptForAuth({ method, endpoint, response, parsed })) {
-      isWriteMethod(method) &&
-      isUnauthorizedResponse(response, parsed) &&
-      basicAuthPromptHandler &&
-      authAttempt < 2
-    ) {
       const credentials = await requestBasicAuthCredentials({
         endpoint,
         method,

orchestrator/src/client/pages/orchestrator/utils.test.ts

@@ -21,6 +21,12 @@ describe("orchestrator utils", () => {
     expect(getEnabledSources(createAppSettings())).toContain("startupjobs");
   });
 
+  it("enables jobicy and themuse without credentials", () => {
+    const enabled = getEnabledSources(createAppSettings());
+    expect(enabled).toContain("jobicy");
+    expect(enabled).toContain("themuse");
+  });
+
   it("counts processing jobs in ready and discovered tabs", () => {
     const jobs = [
       createJob({ id: "ready", status: "ready", closedAt: null }),

orchestrator/src/client/pages/orchestrator/utils.ts

@@ -195,6 +195,21 @@ export const getEnabledSources = (
   const hasAdzunaAuth = Boolean(
     settings.adzunaAppId?.trim() && settings.adzunaAppKeyHint,
   );
+  const hasUsajobsAuth = Boolean(
+    settings.usajobsUserAgent?.trim() && settings.usajobsApiKeyHint,
+  );
+  const hasJoobleAuth = Boolean(settings.joobleApiKeyHint);
+  const hasCareerjetAuth = Boolean(
+    settings.careerjetAffid?.trim() &&
+      settings.careerjetReferer?.trim() &&
+      settings.careerjetUserIp?.trim(),
+  );
+  const hasReedAuth = Boolean(settings.reedApiKeyHint);
+  const hasLeverCompanies = (settings.leverCompanies?.value ?? []).length > 0;
+  const hasAshbyCompanies = (settings.ashbyCompanies?.value ?? []).length > 0;
+  const hasGreenhouseCompanies =
+    (settings.greenhouseCompanies?.value ?? []).length > 0;
+  const hasWorkdayTenants = (settings.workdayTenants?.value ?? []).length > 0;
 
   for (const source of orderedSources) {
     if (source === "gradcracker") {
@@ -209,6 +224,22 @@ export const getEnabledSources = (
       if (hasAdzunaAuth) enabled.push(source);
       continue;
     }
+    if (source === "usajobs") {
+      if (hasUsajobsAuth) enabled.push(source);
+      continue;
+    }
+    if (source === "jooble") {
+      if (hasJoobleAuth) enabled.push(source);
+      continue;
+    }
+    if (source === "careerjet") {
+      if (hasCareerjetAuth) enabled.push(source);
+      continue;
+    }
+    if (source === "reed") {
+      if (hasReedAuth) enabled.push(source);
+      continue;
+    }
     if (source === "hiringcafe") {
       enabled.push(source);
       continue;
@@ -217,10 +248,45 @@ export const getEnabledSources = (
       enabled.push(source);
       continue;
     }
+    if (source === "jobicy") {
+      enabled.push(source);
+      continue;
+    }
+    if (source === "themuse") {
+      enabled.push(source);
+      continue;
+    }
+    if (source === "lever") {
+      if (hasLeverCompanies) enabled.push(source);
+      continue;
+    }
+    if (source === "ashby") {
+      if (hasAshbyCompanies) enabled.push(source);
+      continue;
+    }
+    if (source === "greenhouse") {
+      if (hasGreenhouseCompanies) enabled.push(source);
+      continue;
+    }
+    if (source === "workday") {
+      if (hasWorkdayTenants) enabled.push(source);
+      continue;
+    }
     if (
       source === "indeed" ||
       source === "linkedin" ||
       source === "glassdoor"
+    ) {
+      enabled.push(source);
+      continue;
+    }
+    if (
+      source === "remoteok" ||
+      source === "remotive" ||
+      source === "arbeitnow" ||
+      source === "himalayas" ||
+      source === "weworkremotely" ||
+      source === "fourdayweek"
     ) {
       enabled.push(source);
     }

orchestrator/src/server/api/routes/profiles.ts

@@ -1,4 +1,4 @@
-import { badRequest, forbidden } from "@infra/errors";
+import { badRequest, forbidden, unauthorized } from "@infra/errors";
 import { asyncRoute, fail, ok } from "@infra/http";
 import { logger } from "@infra/logger";
 import {
@@ -28,10 +28,11 @@ function profileMatchesBasicAuthUser(
 function assertProfileVisibleToRequest(
   req: Request,
   profile: SearchProfile,
-): boolean {
+): true | "unauthorized" | "forbidden" {
   if (!isBasicAuthEnabled()) return true;
   const username = parseBasicAuthUsername(req.headers.authorization);
-  return profileMatchesBasicAuthUser(profile, username);
+  if (!username?.trim()) return "unauthorized";
+  return profileMatchesBasicAuthUser(profile, username) ? true : "forbidden";
 }
 
 profilesRouter.get(
@@ -53,7 +54,11 @@ profilesRouter.get(
     if (!profile) {
       return fail(res, badRequest("Profile not found"));
     }
-    if (!assertProfileVisibleToRequest(req, profile)) {
+    const visible = assertProfileVisibleToRequest(req, profile);
+    if (visible === "unauthorized") {
+      return fail(res, unauthorized("Authentication required"));
+    }
+    if (visible === "forbidden") {
       return fail(res, forbidden("You cannot access this profile"));
     }
     return ok(res, profile);
@@ -93,7 +98,11 @@ profilesRouter.patch(
     if (!existing) {
       return fail(res, badRequest("Profile not found"));
     }
-    if (!assertProfileVisibleToRequest(req, existing)) {
+    const visible = assertProfileVisibleToRequest(req, existing);
+    if (visible === "unauthorized") {
+      return fail(res, unauthorized("Authentication required"));
+    }
+    if (visible === "forbidden") {
       return fail(res, forbidden("You cannot update this profile"));
     }
     const { name, data } = req.body;
@@ -139,7 +148,11 @@ profilesRouter.delete(
     if (!existing) {
       return fail(res, badRequest("Profile not found"));
     }
-    if (!assertProfileVisibleToRequest(req, existing)) {
+    const visible = assertProfileVisibleToRequest(req, existing);
+    if (visible === "unauthorized") {
+      return fail(res, unauthorized("Authentication required"));
+    }
+    if (visible === "forbidden") {
       return fail(res, forbidden("You cannot delete this profile"));
     }
     const deleted = await profilesRepo.deleteProfile(req.params.id);
@@ -157,7 +170,11 @@ profilesRouter.post(
     if (!profile) {
       return fail(res, badRequest("Profile not found"));
     }
-    if (!assertProfileVisibleToRequest(req, profile)) {
+    const visible = assertProfileVisibleToRequest(req, profile);
+    if (visible === "unauthorized") {
+      return fail(res, unauthorized("Authentication required"));
+    }
+    if (visible === "forbidden") {
       return fail(res, forbidden("You cannot activate this profile"));
     }
     const basicUser = parseBasicAuthUsername(req.headers.authorization)?.trim();

orchestrator/src/server/api/routes/settings.ts

@@ -10,6 +10,10 @@ import { asyncRoute, fail, ok } from "@infra/http";
 import { logger } from "@infra/logger";
 import { getRequestId } from "@infra/request-context";
 import { isDemoMode, sendDemoBlocked } from "@server/config/demo";
+import { getJobOwnerProfileId } from "@server/infra/request-context";
+import { DEFAULT_JOB_OWNER_PROFILE_ID } from "@server/infra/job-owner-context";
+import { parseBasicAuthUsername } from "@server/infra/basic-auth-credentials";
+import * as profilesRepo from "@server/repositories/profiles";
 import { getSetting } from "@server/repositories/settings";
 import { setBackupSettings } from "@server/services/backup/index";
 import { LlmService } from "@server/services/llm/service";
@@ -30,6 +34,7 @@ import {
   type UpdateSettingsInput,
   updateSettingsSchema,
 } from "@shared/settings-schema";
+import { jobSearchProfileSchema } from "@shared/settings-registry";
 import { type Request, type Response, Router } from "express";
 
 export const settingsRouter = Router();
@@ -232,6 +237,28 @@ settingsRouter.patch(
       clearProfileCache();
     }
 
+    // When Basic Auth is enabled, the effective job search profile comes from
+    // the authenticated user's saved SearchProfile row (basicAuthUser), not the
+    // `jobSearchProfile` setting override. If the client is saving a
+    // jobSearchProfile, persist it directly to the request owner profile so the
+    // next settings fetch immediately reflects the changes (and doesn't "snap back").
+    if (Object.hasOwn(input, "jobSearchProfile") && input.jobSearchProfile) {
+      const ownerId = getJobOwnerProfileId();
+      if (
+        ownerId &&
+        ownerId !== DEFAULT_JOB_OWNER_PROFILE_ID &&
+        ownerId !== "__unmapped__"
+      ) {
+        const parsed = jobSearchProfileSchema.safeParse(input.jobSearchProfile);
+        if (parsed.success) {
+          const username = parseBasicAuthUsername(req.headers.authorization)?.trim();
+          const dataWithOwner =
+            username ? { ...parsed.data, basicAuthUser: username } : parsed.data;
+          await profilesRepo.updateProfile(ownerId, { data: dataWithOwner });
+        }
+      }
+    }
+
     const data = await getEffectiveSettings();
 
     if (plan.shouldRefreshBackupScheduler) {

orchestrator/src/server/config/demo-defaults.data.ts

@@ -256,6 +256,22 @@ export const DEMO_SOURCE_BASE_URLS: Record<JobSource, string> = {
   adzuna: "https://www.adzuna.com",
   hiringcafe: "https://hiring.cafe",
   startupjobs: "https://startup.jobs",
+  usajobs: "https://www.usajobs.gov",
+  jobicy: "https://jobicy.com",
+  themuse: "https://www.themuse.com",
+  jooble: "https://jooble.org",
+  careerjet: "https://www.careerjet.com",
+  reed: "https://www.reed.co.uk",
+  remoteok: "https://remoteok.com",
+  remotive: "https://remotive.com",
+  arbeitnow: "https://www.arbeitnow.com",
+  himalayas: "https://himalayas.app",
+  weworkremotely: "https://weworkremotely.com",
+  fourdayweek: "https://4dayweek.io",
+  ashby: "https://jobs.ashbyhq.com",
+  lever: "https://jobs.lever.co",
+  greenhouse: "https://boards.greenhouse.io",
+  workday: "https://workday.com",
   manual: "https://example.com",
 };
 

orchestrator/src/server/db/migrate.ts

@@ -781,6 +781,13 @@ const migrations = [
   `CREATE INDEX IF NOT EXISTS idx_jobs_status_discovered_at ON jobs(status, discovered_at)`,
   `CREATE INDEX IF NOT EXISTS idx_jobs_owner_profile_id ON jobs(owner_profile_id)`,
 
+  // Cross-source dedup: store a normalized (employer, title) fingerprint and
+  // index it per owner so import-time skip lookups are fast. Backfill happens
+  // lazily in the repository on next insert; existing rows just get NULL until
+  // they're re-imported or rewritten.
+  `ALTER TABLE jobs ADD COLUMN content_fingerprint TEXT`,
+  `CREATE INDEX IF NOT EXISTS idx_jobs_owner_profile_content_fingerprint ON jobs(owner_profile_id, content_fingerprint)`,
+
   // Seed default job-search personas (INSERT OR IGNORE — safe on existing DBs).
   sqlInsertSearchProfileSeed({
     id: "685b0000-0000-4000-8000-000000000001",

orchestrator/src/server/db/schema.ts

@@ -45,6 +45,12 @@ export const jobs = sqliteTable(
     employer: text("employer").notNull(),
     employerUrl: text("employer_url"),
     jobUrl: text("job_url").notNull(),
+    /**
+     * Cross-source dedup key derived from normalized (employer, title).
+     * Nullable because some legacy rows / very thin postings can't produce a
+     * fingerprint. New imports skip when this matches an existing row.
+     */
+    contentFingerprint: text("content_fingerprint"),
     applicationLink: text("application_link"),
     disciplines: text("disciplines"),
     deadline: text("deadline"),
@@ -126,6 +132,9 @@ export const jobs = sqliteTable(
       table.ownerProfileId,
       table.jobUrl,
     ),
+    ownerContentFingerprintIndex: index(
+      "idx_jobs_owner_profile_content_fingerprint",
+    ).on(table.ownerProfileId, table.contentFingerprint),
   }),
 );
 

orchestrator/src/server/pipeline/steps/discover-jobs.ts

@@ -12,6 +12,7 @@ import {
   normalizeCountryKey,
 } from "@shared/location-support.js";
 import { resolveBlockedCompanyKeywordsFromStoredString } from "@shared/resolve-blocked-company-keywords.js";
+import { jobSearchProfileSchema } from "@shared/settings-registry.js";
 import {
   inferCountryKeyFromSearchGeography,
   matchesRequestedCity,
@@ -67,6 +68,127 @@ function filterJobsByRequestedCities(args: {
   );
 }
 
+const ROLE_TOKEN_STOPWORDS = new Set([
+  "a",
+  "an",
+  "and",
+  "the",
+  "of",
+  "to",
+  "for",
+  "in",
+  "on",
+  "with",
+  "at",
+  "by",
+  "from",
+  "senior",
+  "sr",
+  "jr",
+  "junior",
+  "lead",
+  "principal",
+  "staff",
+  "i",
+  "ii",
+  "iii",
+  "iv",
+  "v",
+  "remote",
+  "hybrid",
+  "onsite",
+  // These are too generic and cause massive false positives.
+  "software",
+  "development",
+  "developer",
+  "engineer",
+  "engineering",
+]);
+
+function normalizeText(value: string | null | undefined): string {
+  return (value ?? "")
+    .toLowerCase()
+    .replace(/\s+/g, " ")
+    .trim();
+}
+
+function buildRoleMatchers(phrases: string[]): {
+  phraseMatchers: string[];
+  tokenMatchers: string[];
+} {
+  const phraseMatchers = phrases
+    .map((p) => normalizeText(p))
+    .filter(Boolean);
+
+  const tokenSet = new Set<string>();
+  for (const phrase of phraseMatchers) {
+    for (const token of phrase.split(/[^a-z0-9+.#]+/g)) {
+      const cleaned = token.trim();
+      if (!cleaned) continue;
+      if (cleaned.length < 2) continue;
+      if (ROLE_TOKEN_STOPWORDS.has(cleaned)) continue;
+      tokenSet.add(cleaned);
+    }
+  }
+
+  // Ensure common QA acronyms remain even if user only typed long-form roles.
+  for (const token of ["qa", "sdet", "test", "testing", "automation"]) {
+    tokenSet.add(token);
+  }
+
+  return { phraseMatchers, tokenMatchers: [...tokenSet] };
+}
+
+function matchesAny(text: string, needles: string[]): boolean {
+  if (!text) return false;
+  for (const needle of needles) {
+    if (needle && text.includes(needle)) return true;
+  }
+  return false;
+}
+
+function filterJobsBySearchProfile(args: {
+  jobs: CreateJobInput[];
+  targetRolePhrases: string[];
+  mustHaveSkills: string[];
+  dealBreakers: string[];
+}): { jobs: CreateJobInput[]; dropped: number } {
+  const { jobs, targetRolePhrases, mustHaveSkills, dealBreakers } = args;
+
+  const roleMatchers = buildRoleMatchers(targetRolePhrases);
+  const mustHaveLower = mustHaveSkills.map(normalizeText).filter(Boolean);
+  const dealBreakersLower = dealBreakers.map(normalizeText).filter(Boolean);
+
+  const filtered = jobs.filter((job) => {
+    const title = normalizeText(job.title);
+    const body = normalizeText(job.jobDescription);
+    const haystack = `${title}\n${body}`;
+
+    if (dealBreakersLower.length > 0 && matchesAny(haystack, dealBreakersLower)) {
+      return false;
+    }
+
+    // If the user specified target roles, enforce a strict role match so we
+    // don't surface irrelevant jobs (e.g. legal/sales/finance) in Discovered.
+    if (roleMatchers.phraseMatchers.length > 0) {
+      const roleMatch =
+        matchesAny(title, roleMatchers.phraseMatchers) ||
+        matchesAny(title, roleMatchers.tokenMatchers) ||
+        matchesAny(body, roleMatchers.phraseMatchers) ||
+        matchesAny(body, roleMatchers.tokenMatchers);
+      if (!roleMatch) return false;
+    }
+
+    if (mustHaveLower.length > 0 && !matchesAny(haystack, mustHaveLower)) {
+      return false;
+    }
+
+    return true;
+  });
+
+  return { jobs: filtered, dropped: jobs.length - filtered.length };
+}
+
 export async function discoverJobsStep(args: {
   mergedConfig: PipelineConfig;
   shouldCancel?: () => boolean;
@@ -98,6 +220,9 @@ export async function discoverJobsStep(args: {
 
   const ownerProfileId =
     args.mergedConfig.ownerProfileId ?? DEFAULT_JOB_OWNER_PROFILE_ID;
+  let searchProfileTargetRoles: string[] = [];
+  let searchProfileMustHaveSkills: string[] = [];
+  let searchProfileDealBreakers: string[] = [];
 
   const mergeTargetRoles = (targetRoles: unknown) => {
     if (!Array.isArray(targetRoles) || targetRoles.length === 0) return;
@@ -120,19 +245,38 @@ export async function discoverJobsStep(args: {
 
   if (ownerProfileId && ownerProfileId !== DEFAULT_JOB_OWNER_PROFILE_ID) {
     const row = await getProfileById(ownerProfileId);
-    if (row?.data?.targetRoles?.length) {
+    if (row?.data) {
-      mergeTargetRoles(row.data.targetRoles);
+      const parsed = jobSearchProfileSchema.safeParse(row.data);
+      if (parsed.success) {
+        searchProfileTargetRoles = parsed.data.targetRoles ?? [];
+        searchProfileMustHaveSkills = parsed.data.mustHaveSkills ?? [];
+        searchProfileDealBreakers = parsed.data.dealBreakers ?? [];
+        if (searchProfileTargetRoles.length > 0) {
+          mergeTargetRoles(searchProfileTargetRoles);
+        }
+      } else if (row.data.targetRoles?.length) {
+        // Legacy profile shapes: keep augmenting terms but we won't enforce strict filtering.
+        mergeTargetRoles(row.data.targetRoles);
+      }
     }
   } else {
     const profileSetting = settings.jobSearchProfile;
     if (profileSetting) {
       try {
         const profile = JSON.parse(profileSetting);
-        if (
+        const parsed = jobSearchProfileSchema.safeParse(profile);
-          Array.isArray(profile.targetRoles) &&
+        if (parsed.success) {
-          profile.targetRoles.length > 0
+          searchProfileTargetRoles = parsed.data.targetRoles ?? [];
+          searchProfileMustHaveSkills = parsed.data.mustHaveSkills ?? [];
+          searchProfileDealBreakers = parsed.data.dealBreakers ?? [];
+          if (searchProfileTargetRoles.length > 0) {
+            mergeTargetRoles(searchProfileTargetRoles);
+          }
+        } else if (
+          Array.isArray((profile as { targetRoles?: unknown }).targetRoles) &&
+          (profile as { targetRoles: unknown[] }).targetRoles.length > 0
         ) {
-          mergeTargetRoles(profile.targetRoles);
+          mergeTargetRoles((profile as { targetRoles: unknown }).targetRoles);
         }
       } catch {
         // malformed profile JSON, continue with existing terms
@@ -406,7 +550,32 @@ export async function discoverJobsStep(args: {
     return { discoveredJobs: filteredDiscoveredJobs, sourceErrors };
   }
 
-  if (filteredDiscoveredJobs.length === 0 && sourceErrors.length > 0) {
+  const strictProfileFilteringEnabled =
+    searchProfileTargetRoles.length > 0 ||
+    searchProfileMustHaveSkills.length > 0 ||
+    searchProfileDealBreakers.length > 0;
+  const profileFiltered = strictProfileFilteringEnabled
+    ? filterJobsBySearchProfile({
+        jobs: filteredDiscoveredJobs,
+        targetRolePhrases: searchProfileTargetRoles.length
+          ? searchProfileTargetRoles
+          : searchTerms,
+        mustHaveSkills: searchProfileMustHaveSkills,
+        dealBreakers: searchProfileDealBreakers,
+      })
+    : { jobs: filteredDiscoveredJobs, dropped: 0 };
+
+  if (profileFiltered.dropped > 0) {
+    logger.info("Dropped discovered jobs that didn't match search profile", {
+      step: "discover-jobs",
+      droppedCount: profileFiltered.dropped,
+      targetRolesCount: searchProfileTargetRoles.length,
+      mustHaveSkillsCount: searchProfileMustHaveSkills.length,
+      dealBreakersCount: searchProfileDealBreakers.length,
+    });
+  }
+
+  if (profileFiltered.jobs.length === 0 && sourceErrors.length > 0) {
     throw new Error(`All sources failed: ${sourceErrors.join("; ")}`);
   }
 
@@ -414,9 +583,9 @@ export async function discoverJobsStep(args: {
     logger.warn("Some discovery sources failed", { sourceErrors });
   }
 
-  progressHelpers.crawlingComplete(filteredDiscoveredJobs.length);
+  progressHelpers.crawlingComplete(profileFiltered.jobs.length);
 
-  const stamped = filteredDiscoveredJobs.map((job) => ({
+  const stamped = profileFiltered.jobs.map((job) => ({
     ...job,
     ownerProfileId,
   }));

orchestrator/src/server/repositories/jobs.ts

@@ -5,6 +5,7 @@
 import { randomUUID } from "node:crypto";
 import { getJobOwnerProfileId } from "@infra/request-context";
 import { DEFAULT_JOB_OWNER_PROFILE_ID } from "@server/infra/job-owner-context";
+import { buildJobContentFingerprint } from "@shared/job-fingerprint";
 import { canonicalizeJobUrl } from "@shared/job-url-canonical";
 import type {
   CreateJobInput,
@@ -40,12 +41,16 @@ function resolveOwnerForCreate(input: CreateJobInput): string {
 async function loadJobDedupIndexes(ownerProfileId: string): Promise<{
   existingCanonicalSet: Set<string>;
   existingSourceJobKeySet: Set<string>;
+  existingContentFingerprintSet: Set<string>;
 }> {
   const rows = await db
     .select({
       jobUrl: jobs.jobUrl,
       source: jobs.source,
       sourceJobId: jobs.sourceJobId,
+      contentFingerprint: jobs.contentFingerprint,
+      employer: jobs.employer,
+      title: jobs.title,
     })
     .from(jobs)
     .where(eq(jobs.ownerProfileId, ownerProfileId));
@@ -60,7 +65,29 @@ async function loadJobDedupIndexes(ownerProfileId: string): Promise<{
       )
       .map((r) => sourceJobKey(r.source, String(r.sourceJobId))),
   );
-  return { existingCanonicalSet, existingSourceJobKeySet };
+  // Cross-source dedup: prefer the persisted fingerprint, but fall back to
+  // recomputing it from (employer, title) so legacy rows participate in
+  // dedup until they're rewritten.
+  const existingContentFingerprintSet = new Set<string>();
+  for (const row of rows) {
+    const stored = row.contentFingerprint?.trim();
+    if (stored) {
+      existingContentFingerprintSet.add(stored);
+      continue;
+    }
+    const recomputed = buildJobContentFingerprint({
+      employer: row.employer,
+      title: row.title,
+    });
+    if (recomputed) {
+      existingContentFingerprintSet.add(recomputed);
+    }
+  }
+  return {
+    existingCanonicalSet,
+    existingSourceJobKeySet,
+    existingContentFingerprintSet,
+  };
 }
 
 async function findJobByCanonicalUrl(
@@ -87,6 +114,46 @@ async function findJobByCanonicalUrl(
   return null;
 }
 
+async function findJobByContentFingerprint(
+  fingerprint: string,
+  ownerProfileId: string,
+): Promise<Job | null> {
+  // Fast path: stored fingerprint match.
+  const [stored] = await db
+    .select()
+    .from(jobs)
+    .where(
+      and(
+        eq(jobs.ownerProfileId, ownerProfileId),
+        eq(jobs.contentFingerprint, fingerprint),
+      ),
+    )
+    .limit(1);
+  if (stored) return mapRowToJob(stored);
+
+  // Fallback for legacy rows without a persisted fingerprint: scan and
+  // recompute. Owner-scoped table size keeps this cheap in practice.
+  const allRows = await db
+    .select()
+    .from(jobs)
+    .where(
+      and(
+        eq(jobs.ownerProfileId, ownerProfileId),
+        isNull(jobs.contentFingerprint),
+      ),
+    );
+  for (const row of allRows) {
+    const recomputed = buildJobContentFingerprint({
+      employer: row.employer,
+      title: row.title,
+    });
+    if (recomputed === fingerprint) {
+      return mapRowToJob(row);
+    }
+  }
+  return null;
+}
+
 async function getJobBySourceAndExternalId(
   source: string,
   sourceJobId: string,
@@ -302,6 +369,10 @@ async function insertJob(input: CreateJobInput): Promise<Job> {
   const now = new Date().toISOString();
 
   const ownerProfileId = resolveOwnerForCreate(input);
+  const contentFingerprint = buildJobContentFingerprint({
+    employer: input.employer,
+    title: input.title,
+  });
 
   await db.insert(jobs).values({
     id,
@@ -314,6 +385,7 @@ async function insertJob(input: CreateJobInput): Promise<Job> {
     employer: input.employer,
     employerUrl: input.employerUrl ?? null,
     jobUrl: input.jobUrl,
+    contentFingerprint,
     applicationLink: input.applicationLink ?? null,
     disciplines: input.disciplines ?? null,
     deadline: input.deadline ?? null,
@@ -395,8 +467,11 @@ export async function createJobs(
       ...normalized,
       ownerProfileId,
     };
-    const { existingCanonicalSet, existingSourceJobKeySet } =
+    const {
-      await loadJobDedupIndexes(ownerProfileId);
+      existingCanonicalSet,
+      existingSourceJobKeySet,
+      existingContentFingerprintSet,
+    } = await loadJobDedupIndexes(ownerProfileId);
 
     const sid = normalized.sourceJobId?.trim();
     if (sid) {
@@ -419,6 +494,18 @@ export async function createJobs(
       if (existing) return existing;
     }
 
+    const fingerprint = buildJobContentFingerprint({
+      employer: normalized.employer,
+      title: normalized.title,
+    });
+    if (fingerprint && existingContentFingerprintSet.has(fingerprint)) {
+      const existing = await findJobByContentFingerprint(
+        fingerprint,
+        ownerProfileId,
+      );
+      if (existing) return existing;
+    }
+
     const inserted = await tryInsertJob(normalizedWithOwner);
     if (inserted) return inserted;
 
@@ -437,14 +524,18 @@ export async function createJobs(
   }
 
   const ownerProfileId = resolveOwnerForCreate(inputOrInputs[0] ?? {});
-  const { existingCanonicalSet, existingSourceJobKeySet } =
+  const {
-    await loadJobDedupIndexes(ownerProfileId);
+    existingCanonicalSet,
+    existingSourceJobKeySet,
+    existingContentFingerprintSet,
+  } = await loadJobDedupIndexes(ownerProfileId);
 
   const batchBuckets = new Map<
     string,
     {
       input: CreateJobInput;
       count: number;
+      fingerprint: string | null;
     }
   >();
 
@@ -454,21 +545,30 @@ export async function createJobs(
       ownerProfileId,
     });
     const sidForKey = normalized.sourceJobId?.trim();
-    const batchKey = sidForKey
+    const fingerprint = buildJobContentFingerprint({
-      ? `sid:${sourceJobKey(normalized.source, sidForKey)}`
+      employer: normalized.employer,
-      : `url:${normalized.jobUrl}`;
+      title: normalized.title,
+    });
+    // Coalesce duplicates within a single batch, preferring fingerprint when
+    // available so two different feeds posting the same role merge into one
+    // bucket. Fall back to source-job-id, then canonical URL.
+    const batchKey = fingerprint
+      ? `fp:${fingerprint}`
+      : sidForKey
+        ? `sid:${sourceJobKey(normalized.source, sidForKey)}`
+        : `url:${normalized.jobUrl}`;
     const prev = batchBuckets.get(batchKey);
     if (prev) {
       prev.count += 1;
     } else {
-      batchBuckets.set(batchKey, { input: normalized, count: 1 });
+      batchBuckets.set(batchKey, { input: normalized, count: 1, fingerprint });
     }
   }
 
   let created = 0;
   let skipped = 0;
 
-  for (const { input, count } of batchBuckets.values()) {
+  for (const { input, count, fingerprint } of batchBuckets.values()) {
     const canonical = input.jobUrl;
     const sid = input.sourceJobId?.trim();
     const sk = sid ? sourceJobKey(input.source, sid) : null;
@@ -481,6 +581,10 @@ export async function createJobs(
       skipped += count;
       continue;
     }
+    if (fingerprint && existingContentFingerprintSet.has(fingerprint)) {
+      skipped += count;
+      continue;
+    }
 
     const inserted = await tryInsertJob(input);
     if (!inserted) {
@@ -494,6 +598,9 @@ export async function createJobs(
     if (sk) {
       existingSourceJobKeySet.add(sk);
     }
+    if (fingerprint) {
+      existingContentFingerprintSet.add(fingerprint);
+    }
   }
 
   return { created, skipped };

orchestrator/src/server/services/settings.ts

@@ -151,6 +151,20 @@ export async function getEffectiveSettings(): Promise<AppSettings> {
     ...envSettings,
   };
 
+  // In Basic Auth mode, the "active" search profile is derived from the request
+  // context (basic auth username → profile.basicAuthUser), not from the stored
+  // activeProfileId setting. Expose that derived id so the UI updates the right
+  // profile record on save.
+  const requestOwnerProfileId = getJobOwnerProfileId();
+  if (
+    tenantJobSearchProfile &&
+    requestOwnerProfileId &&
+    requestOwnerProfileId !== DEFAULT_JOB_OWNER_PROFILE_ID &&
+    requestOwnerProfileId !== "__unmapped__"
+  ) {
+    result.activeProfileId = requestOwnerProfileId;
+  }
+
   const rawModel = overrides.model;
   const modelDef = settingsRegistry.model;
   const overrideModel = normalizeModelForProviderCompatibility(

package-lock.json

@@ -13,6 +13,7 @@
       ],
       "devDependencies": {
         "@types/node": "^25.2.3",
+        "dotenv": "^17.2.3",
         "knip": "^5.83.1",
         "tsx": "^4.19.2",
         "typescript": "^5.9.3"
@@ -103,6 +104,90 @@
         "undici-types": "~7.16.0"
       }
     },
+    "extractors/arbeitnow": {
+      "name": "arbeitnow-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/arbeitnow/node_modules/@types/node": {
+      "version": "24.12.4",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.4.tgz",
+      "integrity": "sha512-GUUEShf+PBCGW2KaXwcIt3Yk+e3pkKwWKb9GSyM9WQVE+ep2jzmHdGsHzu4wgcZy5fN9FBdVzjpBQsYlpfpgLA==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
+    "extractors/ashby": {
+      "name": "ashby-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/ashby/node_modules/@types/node": {
+      "version": "24.12.3",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.3.tgz",
+      "integrity": "sha512-8oljBDGun9cIsZRJR6fkihn0TSXJI0UDOOhncYaERq6M0JMDoPLxyscwruJcb4GKS6dvK/d8xebYBg27h/duaQ==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
+    "extractors/careerjet": {
+      "name": "careerjet-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/careerjet/node_modules/@types/node": {
+      "version": "24.12.3",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.3.tgz",
+      "integrity": "sha512-8oljBDGun9cIsZRJR6fkihn0TSXJI0UDOOhncYaERq6M0JMDoPLxyscwruJcb4GKS6dvK/d8xebYBg27h/duaQ==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
+    "extractors/fourdayweek": {
+      "name": "fourdayweek-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/fourdayweek/node_modules/@types/node": {
+      "version": "24.12.4",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.4.tgz",
+      "integrity": "sha512-GUUEShf+PBCGW2KaXwcIt3Yk+e3pkKwWKb9GSyM9WQVE+ep2jzmHdGsHzu4wgcZy5fN9FBdVzjpBQsYlpfpgLA==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
     "extractors/gradcracker": {
       "name": "gradcracker-extractor",
       "version": "0.0.1",
@@ -154,6 +239,48 @@
         "undici-types": "~7.16.0"
       }
     },
+    "extractors/greenhouse": {
+      "name": "greenhouse-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/greenhouse/node_modules/@types/node": {
+      "version": "24.12.3",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.3.tgz",
+      "integrity": "sha512-8oljBDGun9cIsZRJR6fkihn0TSXJI0UDOOhncYaERq6M0JMDoPLxyscwruJcb4GKS6dvK/d8xebYBg27h/duaQ==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
+    "extractors/himalayas": {
+      "name": "himalayas-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/himalayas/node_modules/@types/node": {
+      "version": "24.12.4",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.4.tgz",
+      "integrity": "sha512-GUUEShf+PBCGW2KaXwcIt3Yk+e3pkKwWKb9GSyM9WQVE+ep2jzmHdGsHzu4wgcZy5fN9FBdVzjpBQsYlpfpgLA==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
     "extractors/hiringcafe": {
       "name": "hiringcafe-extractor",
       "version": "0.0.1",
@@ -181,6 +308,132 @@
         "undici-types": "~7.16.0"
       }
     },
+    "extractors/jobicy": {
+      "name": "jobicy-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/jobicy/node_modules/@types/node": {
+      "version": "24.12.3",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.3.tgz",
+      "integrity": "sha512-8oljBDGun9cIsZRJR6fkihn0TSXJI0UDOOhncYaERq6M0JMDoPLxyscwruJcb4GKS6dvK/d8xebYBg27h/duaQ==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
+    "extractors/jooble": {
+      "name": "jooble-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/jooble/node_modules/@types/node": {
+      "version": "24.12.3",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.3.tgz",
+      "integrity": "sha512-8oljBDGun9cIsZRJR6fkihn0TSXJI0UDOOhncYaERq6M0JMDoPLxyscwruJcb4GKS6dvK/d8xebYBg27h/duaQ==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
+    "extractors/lever": {
+      "name": "lever-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/lever/node_modules/@types/node": {
+      "version": "24.12.3",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.3.tgz",
+      "integrity": "sha512-8oljBDGun9cIsZRJR6fkihn0TSXJI0UDOOhncYaERq6M0JMDoPLxyscwruJcb4GKS6dvK/d8xebYBg27h/duaQ==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
+    "extractors/reed": {
+      "name": "reed-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/reed/node_modules/@types/node": {
+      "version": "24.12.3",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.3.tgz",
+      "integrity": "sha512-8oljBDGun9cIsZRJR6fkihn0TSXJI0UDOOhncYaERq6M0JMDoPLxyscwruJcb4GKS6dvK/d8xebYBg27h/duaQ==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
+    "extractors/remoteok": {
+      "name": "remoteok-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/remoteok/node_modules/@types/node": {
+      "version": "24.12.3",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.3.tgz",
+      "integrity": "sha512-8oljBDGun9cIsZRJR6fkihn0TSXJI0UDOOhncYaERq6M0JMDoPLxyscwruJcb4GKS6dvK/d8xebYBg27h/duaQ==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
+    "extractors/remotive": {
+      "name": "remotive-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/remotive/node_modules/@types/node": {
+      "version": "24.12.4",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.4.tgz",
+      "integrity": "sha512-GUUEShf+PBCGW2KaXwcIt3Yk+e3pkKwWKb9GSyM9WQVE+ep2jzmHdGsHzu4wgcZy5fN9FBdVzjpBQsYlpfpgLA==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
     "extractors/startupjobs": {
       "name": "startupjobs-extractor",
       "version": "0.0.1",
@@ -202,6 +455,27 @@
         "undici-types": "~7.16.0"
       }
     },
+    "extractors/themuse": {
+      "name": "themuse-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/themuse/node_modules/@types/node": {
+      "version": "24.12.3",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.3.tgz",
+      "integrity": "sha512-8oljBDGun9cIsZRJR6fkihn0TSXJI0UDOOhncYaERq6M0JMDoPLxyscwruJcb4GKS6dvK/d8xebYBg27h/duaQ==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
     "extractors/ukvisajobs": {
       "name": "ukvisajobs-extractor",
       "version": "0.0.1",
@@ -234,6 +508,69 @@
         "undici-types": "~7.16.0"
       }
     },
+    "extractors/usajobs": {
+      "name": "usajobs-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/usajobs/node_modules/@types/node": {
+      "version": "24.12.3",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.3.tgz",
+      "integrity": "sha512-8oljBDGun9cIsZRJR6fkihn0TSXJI0UDOOhncYaERq6M0JMDoPLxyscwruJcb4GKS6dvK/d8xebYBg27h/duaQ==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
+    "extractors/weworkremotely": {
+      "name": "weworkremotely-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/weworkremotely/node_modules/@types/node": {
+      "version": "24.12.4",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.4.tgz",
+      "integrity": "sha512-GUUEShf+PBCGW2KaXwcIt3Yk+e3pkKwWKb9GSyM9WQVE+ep2jzmHdGsHzu4wgcZy5fN9FBdVzjpBQsYlpfpgLA==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
+    "extractors/workday": {
+      "name": "workday-extractor",
+      "version": "0.0.1",
+      "dependencies": {
+        "job-ops-shared": "^1.0.0"
+      },
+      "devDependencies": {
+        "@types/node": "^24.0.0",
+        "typescript": "~5.9.0"
+      }
+    },
+    "extractors/workday/node_modules/@types/node": {
+      "version": "24.12.3",
+      "resolved": "https://registry.npmjs.org/@types/node/-/node-24.12.3.tgz",
+      "integrity": "sha512-8oljBDGun9cIsZRJR6fkihn0TSXJI0UDOOhncYaERq6M0JMDoPLxyscwruJcb4GKS6dvK/d8xebYBg27h/duaQ==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "undici-types": "~7.16.0"
+      }
+    },
     "node_modules/@algolia/abtesting": {
       "version": "1.14.1",
       "resolved": "https://registry.npmjs.org/@algolia/abtesting/-/abtesting-1.14.1.tgz",
@@ -9384,6 +9721,10 @@
         "type-fest": "^4.0.0"
       }
     },
+    "node_modules/arbeitnow-extractor": {
+      "resolved": "extractors/arbeitnow",
+      "link": true
+    },
     "node_modules/arg": {
       "version": "5.0.2",
       "resolved": "https://registry.npmjs.org/arg/-/arg-5.0.2.tgz",
@@ -9423,6 +9764,10 @@
         "node": ">=8"
       }
     },
+    "node_modules/ashby-extractor": {
+      "resolved": "extractors/ashby",
+      "link": true
+    },
     "node_modules/asn1js": {
       "version": "3.0.7",
       "resolved": "https://registry.npmjs.org/asn1js/-/asn1js-3.0.7.tgz",
@@ -10208,6 +10553,10 @@
       ],
       "license": "CC-BY-4.0"
     },
+    "node_modules/careerjet-extractor": {
+      "resolved": "extractors/careerjet",
+      "link": true
+    },
     "node_modules/ccount": {
       "version": "2.0.1",
       "resolved": "https://registry.npmjs.org/ccount/-/ccount-2.0.1.tgz",
@@ -11899,6 +12248,18 @@
         "url": "https://github.com/sponsors/sindresorhus"
       }
     },
+    "node_modules/dotenv": {
+      "version": "17.4.2",
+      "resolved": "https://registry.npmjs.org/dotenv/-/dotenv-17.4.2.tgz",
+      "integrity": "sha512-nI4U3TottKAcAD9LLud4Cb7b2QztQMUEfHbvhTH09bqXTxnSie8WnjPALV/WMCrJZ6UV/qHJ6L03OqO3LcdYZw==",
+      "license": "BSD-2-Clause",
+      "engines": {
+        "node": ">=12"
+      },
+      "funding": {
+        "url": "https://dotenvx.com"
+      }
+    },
     "node_modules/dunder-proto": {
       "version": "1.0.1",
       "resolved": "https://registry.npmjs.org/dunder-proto/-/dunder-proto-1.0.1.tgz",
@@ -13168,6 +13529,10 @@
         "node": ">= 0.6"
       }
     },
+    "node_modules/fourdayweek-extractor": {
+      "resolved": "extractors/fourdayweek",
+      "link": true
+    },
     "node_modules/fraction.js": {
       "version": "5.3.4",
       "resolved": "https://registry.npmjs.org/fraction.js/-/fraction.js-5.3.4.tgz",
@@ -13695,6 +14060,10 @@
         "js-yaml": "bin/js-yaml.js"
       }
     },
+    "node_modules/greenhouse-extractor": {
+      "resolved": "extractors/greenhouse",
+      "link": true
+    },
     "node_modules/gzip-size": {
       "version": "6.0.0",
       "resolved": "https://registry.npmjs.org/gzip-size/-/gzip-size-6.0.0.tgz",
@@ -13995,6 +14364,10 @@
         "node": ">=16.0.0"
       }
     },
+    "node_modules/himalayas-extractor": {
+      "resolved": "extractors/himalayas",
+      "link": true
+    },
     "node_modules/hiringcafe-extractor": {
       "resolved": "extractors/hiringcafe",
       "link": true
@@ -15261,6 +15634,10 @@
       "resolved": "shared",
       "link": true
     },
+    "node_modules/jobicy-extractor": {
+      "resolved": "extractors/jobicy",
+      "link": true
+    },
     "node_modules/joi": {
       "version": "17.13.3",
       "resolved": "https://registry.npmjs.org/joi/-/joi-17.13.3.tgz",
@@ -15274,6 +15651,10 @@
         "@sideway/pinpoint": "^2.0.0"
       }
     },
+    "node_modules/jooble-extractor": {
+      "resolved": "extractors/jooble",
+      "link": true
+    },
     "node_modules/jquery": {
       "version": "3.7.1",
       "resolved": "https://registry.npmjs.org/jquery/-/jquery-3.7.1.tgz",
@@ -15573,6 +15954,10 @@
         "node": ">=6"
       }
     },
+    "node_modules/lever-extractor": {
+      "resolved": "extractors/lever",
+      "link": true
+    },
     "node_modules/lightningcss-android-arm64": {
       "version": "1.30.2",
       "resolved": "https://registry.npmjs.org/lightningcss-android-arm64/-/lightningcss-android-arm64-1.30.2.tgz",
@@ -21836,6 +22221,10 @@
         "url": "https://opencollective.com/unified"
       }
     },
+    "node_modules/reed-extractor": {
+      "resolved": "extractors/reed",
+      "link": true
+    },
     "node_modules/reflect-metadata": {
       "version": "0.2.2",
       "resolved": "https://registry.npmjs.org/reflect-metadata/-/reflect-metadata-0.2.2.tgz",
@@ -22089,6 +22478,14 @@
         "url": "https://opencollective.com/unified"
       }
     },
+    "node_modules/remoteok-extractor": {
+      "resolved": "extractors/remoteok",
+      "link": true
+    },
+    "node_modules/remotive-extractor": {
+      "resolved": "extractors/remotive",
+      "link": true
+    },
     "node_modules/renderkid": {
       "version": "3.0.0",
       "resolved": "https://registry.npmjs.org/renderkid/-/renderkid-3.0.0.tgz",
@@ -23752,6 +24149,10 @@
       "integrity": "sha512-GpVkmM8vF2vQUkj2LvZmD35JxeJOLCwJ9cUkugyk2nuhbv3+mJvpLYYt+0+USMxE+oj+ey/lJEnhZw75x/OMcQ==",
       "license": "MIT"
     },
+    "node_modules/themuse-extractor": {
+      "resolved": "extractors/themuse",
+      "link": true
+    },
     "node_modules/thingies": {
       "version": "2.5.0",
       "resolved": "https://registry.npmjs.org/thingies/-/thingies-2.5.0.tgz",
@@ -24601,6 +25002,10 @@
         "url": "https://opencollective.com/webpack"
       }
     },
+    "node_modules/usajobs-extractor": {
+      "resolved": "extractors/usajobs",
+      "link": true
+    },
     "node_modules/use-callback-ref": {
       "version": "1.3.3",
       "resolved": "https://registry.npmjs.org/use-callback-ref/-/use-callback-ref-1.3.3.tgz",
@@ -25191,6 +25596,10 @@
         "node": ">=0.8.0"
       }
     },
+    "node_modules/weworkremotely-extractor": {
+      "resolved": "extractors/weworkremotely",
+      "link": true
+    },
     "node_modules/whatwg-encoding": {
       "version": "3.1.1",
       "resolved": "https://registry.npmjs.org/whatwg-encoding/-/whatwg-encoding-3.1.1.tgz",
@@ -25324,6 +25733,10 @@
       "integrity": "sha512-CC1bOL87PIWSBhDcTrdeLo6eGT7mCFtrg0uIJtqJUFyK+eJnzl8A1niH56uu7KMa5XFrtiV+AQuHO3n7DsHnLQ==",
       "license": "MIT"
     },
+    "node_modules/workday-extractor": {
+      "resolved": "extractors/workday",
+      "link": true
+    },
     "node_modules/wrap-ansi": {
       "version": "6.2.0",
       "resolved": "https://registry.npmjs.org/wrap-ansi/-/wrap-ansi-6.2.0.tgz",
@@ -25640,7 +26053,7 @@
     },
     "orchestrator": {
       "name": "job-ops-orchestrator",
-      "version": "0.2.0",
+      "version": "0.2.1",
       "dependencies": {
         "@hookform/resolvers": "^5.2.2",
         "@paralleldrive/cuid2": "^3.0.6",
@@ -27412,16 +27825,6 @@
         "csstype": "^3.0.2"
       }
     },
-    "orchestrator/node_modules/dotenv": {
-      "version": "17.2.3",
-      "license": "BSD-2-Clause",
-      "engines": {
-        "node": ">=12"
-      },
-      "funding": {
-        "url": "https://dotenvx.com"
-      }
-    },
     "orchestrator/node_modules/drizzle-kit": {
       "version": "0.30.6",
       "dev": true,

package.json

@@ -24,6 +24,7 @@
     "knip": "knip"
   },
   "devDependencies": {
+    "dotenv": "^17.2.3",
     "@types/node": "^25.2.3",
     "knip": "^5.83.1",
     "tsx": "^4.19.2",

scripts/jobber-cron-cherepaha.env.example

@@ -0,0 +1,24 @@
+#
+# Jobber cron env — cherepaha
+# Copy to /root/.jobber-cron-cherepaha.env (chmod 600)
+#
+# Used by: scripts/jobber-pipeline-telegram.sh
+#
+
+# JobOps base URL (where the app is reachable from the cron host)
+JOBOPS_URL=http://127.0.0.1:3005
+
+# Optional: limit number of jobs linked in Telegram message
+JOB_TELEGRAM_MAX_JOBS=25
+
+# Optional: comma-separated sources to run (leave empty to use server defaults)
+# JOBBER_PIPELINE_SOURCES=adzuna,gradcracker,ukvisajobs
+
+# App-level Basic Auth (enables per-user separation when set on the server)
+BASIC_AUTH_USER=cherepaha
+BASIC_AUTH_PASSWORD=CHANGEME
+
+# Telegram bot + chat destination
+TELEGRAM_BOT_TOKEN=CHANGEME
+TELEGRAM_CHAT_ID=CHANGEME
+

scripts/jobber-cron-dobkin.env.example

@@ -0,0 +1,24 @@
+#
+# Jobber cron env — dobkin
+# Copy to /root/.jobber-cron-dobkin.env (chmod 600)
+#
+# Used by: scripts/jobber-pipeline-telegram.sh
+#
+
+# JobOps base URL (where the app is reachable from the cron host)
+JOBOPS_URL=http://127.0.0.1:3005
+
+# Optional: limit number of jobs linked in Telegram message
+JOB_TELEGRAM_MAX_JOBS=25
+
+# Optional: comma-separated sources to run (leave empty to use server defaults)
+# JOBBER_PIPELINE_SOURCES=adzuna,gradcracker,ukvisajobs
+
+# App-level Basic Auth (enables per-user separation when set on the server)
+BASIC_AUTH_USER=dobkin
+BASIC_AUTH_PASSWORD=CHANGEME
+
+# Telegram bot + chat destination
+TELEGRAM_BOT_TOKEN=CHANGEME
+TELEGRAM_CHAT_ID=CHANGEME
+

scripts/smoke-extractors.ts

@@ -0,0 +1,219 @@
+/**
+ * Tiny smoke-test for new extractors: imports each manifest, runs it with a
+ * minimal context, and prints the count of mapped jobs + a few samples.
+ *
+ * Run from repo root:  npx tsx scripts/smoke-extractors.ts [comma,separated,ids]
+ *
+ * Loads repo-root `.env` so keyed extractors match orchestrator behavior (plain
+ * `tsx` does not read `.env` automatically).
+ */
+
+import path from "node:path";
+import { fileURLToPath } from "node:url";
+import { config as loadEnv } from "dotenv";
+import type {
+  ExtractorManifest,
+  ExtractorRuntimeContext,
+} from "../shared/src/types/extractors";
+
+const repoRoot = path.resolve(
+  path.dirname(fileURLToPath(import.meta.url)),
+  "..",
+);
+loadEnv({ path: path.join(repoRoot, ".env") });
+
+interface Target {
+  id: string;
+  importPath: string;
+  needs?: string[]; // env vars required to run; skipped if missing
+  settings?: Record<string, string>;
+}
+
+const ALL_TARGETS: Target[] = [
+  {
+    id: "jobicy",
+    importPath: "../extractors/jobicy/manifest",
+    settings: { jobicyMaxJobsPerTerm: "10" },
+  },
+  {
+    id: "themuse",
+    importPath: "../extractors/themuse/manifest",
+    settings: { themuseMaxJobsPerTerm: "10" },
+  },
+  {
+    id: "usajobs",
+    importPath: "../extractors/usajobs/manifest",
+    needs: ["USAJOBS_API_KEY", "USAJOBS_USER_AGENT"],
+    settings: { usajobsMaxJobsPerTerm: "10" },
+  },
+  {
+    id: "jooble",
+    importPath: "../extractors/jooble/manifest",
+    needs: ["JOOBLE_API_KEY"],
+    settings: { joobleMaxJobsPerTerm: "10" },
+  },
+  {
+    id: "careerjet",
+    importPath: "../extractors/careerjet/manifest",
+    needs: ["CAREERJET_AFFID", "CAREERJET_REFERER", "CAREERJET_USER_IP"],
+    settings: { careerjetMaxJobsPerTerm: "10" },
+  },
+  {
+    id: "reed",
+    importPath: "../extractors/reed/manifest",
+    needs: ["REED_API_KEY"],
+    settings: { reedMaxJobsPerTerm: "10" },
+  },
+  {
+    id: "lever",
+    importPath: "../extractors/lever/manifest",
+    settings: {
+      // Known active public Lever board used purely as a connectivity check.
+      leverCompanies: JSON.stringify(["palantir", "netflix"]),
+    },
+  },
+  {
+    id: "ashby",
+    importPath: "../extractors/ashby/manifest",
+    settings: {
+      ashbyCompanies: JSON.stringify(["ramp", "linear"]),
+    },
+  },
+  {
+    id: "greenhouse",
+    importPath: "../extractors/greenhouse/manifest",
+    settings: {
+      greenhouseCompanies: JSON.stringify(["stripe", "airbnb"]),
+    },
+  },
+  {
+    id: "workday",
+    importPath: "../extractors/workday/manifest",
+    settings: {
+      workdayTenants: JSON.stringify([
+        "https://nvidia.wd5.myworkdayjobs.com/en-US/NVIDIAExternalCareerSite",
+      ]),
+    },
+  },
+  {
+    id: "remoteok",
+    importPath: "../extractors/remoteok/manifest",
+    settings: { remoteokMaxJobsPerTerm: "10" },
+  },
+  {
+    id: "remotive",
+    importPath: "../extractors/remotive/manifest",
+    settings: { remotiveMaxJobsPerTerm: "10" },
+  },
+  {
+    id: "arbeitnow",
+    importPath: "../extractors/arbeitnow/manifest",
+    settings: { arbeitnowMaxJobsPerTerm: "10" },
+  },
+  {
+    id: "himalayas",
+    importPath: "../extractors/himalayas/manifest",
+    settings: { himalayasMaxJobsPerTerm: "10" },
+  },
+  {
+    id: "weworkremotely",
+    importPath: "../extractors/weworkremotely/manifest",
+    settings: { weworkremotelyMaxJobsPerTerm: "10" },
+  },
+  {
+    id: "fourdayweek",
+    importPath: "../extractors/fourdayweek/manifest",
+    settings: { fourdayweekMaxJobsPerTerm: "10" },
+  },
+];
+
+function buildContext(
+  source: string,
+  settings: Record<string, string>,
+): ExtractorRuntimeContext {
+  return {
+    source,
+    selectedSources: [source],
+    settings,
+    searchTerms: ["software engineer"],
+    selectedCountry: "United States",
+    getExistingJobUrls: async () => [],
+    shouldCancel: () => false,
+    onProgress: () => {},
+  };
+}
+
+function pad(s: string, n: number): string {
+  return s.length >= n ? s : s + " ".repeat(n - s.length);
+}
+
+async function runOne(target: Target): Promise<void> {
+  const missing = (target.needs ?? []).filter((k) => !process.env[k]);
+  if (missing.length > 0) {
+    console.log(
+      `${pad(target.id, 12)}  SKIP  missing env: ${missing.join(", ")}`,
+    );
+    return;
+  }
+
+  let mod: { manifest?: ExtractorManifest; default?: ExtractorManifest };
+  try {
+    mod = await import(target.importPath);
+  } catch (err) {
+    console.log(
+      `${pad(target.id, 12)}  FAIL  import error: ${(err as Error).message}`,
+    );
+    return;
+  }
+
+  const manifest = mod.manifest ?? mod.default;
+  if (!manifest) {
+    console.log(`${pad(target.id, 12)}  FAIL  manifest export missing`);
+    return;
+  }
+
+  const started = Date.now();
+  try {
+    const ctx = buildContext(target.id, target.settings ?? {});
+    const result = await manifest.run(ctx);
+    const ms = Date.now() - started;
+    const status = result.success ? "OK  " : "ERR ";
+    const sample = result.jobs[0];
+    const sampleStr = sample
+      ? ` | first: "${sample.title}" @ ${sample.employer}`
+      : "";
+    console.log(
+      `${pad(target.id, 12)}  ${status} jobs=${result.jobs.length}  ${ms}ms${result.error ? ` | error: ${result.error}` : ""}${sampleStr}`,
+    );
+  } catch (err) {
+    const ms = Date.now() - started;
+    console.log(
+      `${pad(target.id, 12)}  CRASH ${ms}ms  ${(err as Error).message}`,
+    );
+  }
+}
+
+async function main() {
+  const requested = (process.argv[2] ?? "").trim();
+  const filter = requested
+    ? new Set(
+        requested
+          .split(",")
+          .map((s) => s.trim())
+          .filter(Boolean),
+      )
+    : null;
+  const targets = filter
+    ? ALL_TARGETS.filter((t) => filter.has(t.id))
+    : ALL_TARGETS;
+
+  console.log(`Smoke testing ${targets.length} extractor(s)...\n`);
+  for (const t of targets) {
+    await runOne(t);
+  }
+}
+
+main().catch((err) => {
+  console.error(err);
+  process.exit(1);
+});

shared/src/extractors/index.ts

@@ -9,6 +9,24 @@ export const EXTRACTOR_SOURCE_IDS = [
   "adzuna",
   "hiringcafe",
   "startupjobs",
+  // --- Public APIs / feeds ---
+  "usajobs",
+  "jobicy",
+  "themuse",
+  "jooble",
+  "careerjet",
+  "reed",
+  "remoteok",
+  "remotive",
+  "arbeitnow",
+  "himalayas",
+  "weworkremotely",
+  "fourdayweek",
+  // --- Public ATS / career-page sources ---
+  "ashby",
+  "lever",
+  "greenhouse",
+  "workday",
   "manual",
 ] as const;
 
@@ -20,6 +38,10 @@ export interface ExtractorSourceMetadata {
   category: "pipeline" | "manual";
   requiresCredentials?: boolean;
   ukOnly?: boolean;
+  /** Country gating: when set, only run/show this source for these country keys. */
+  countryAllowlist?: readonly string[];
+  /** Region tag for grouping / filtering in the UI. */
+  region?: "us" | "uk" | "global" | "remote";
 }
 
 export const EXTRACTOR_SOURCE_METADATA: Record<
@@ -31,26 +53,157 @@ export const EXTRACTOR_SOURCE_METADATA: Record<
     order: 10,
     category: "pipeline",
     ukOnly: true,
+    region: "uk",
+  },
+  indeed: {
+    label: "Indeed",
+    order: 20,
+    category: "pipeline",
+    region: "global",
+  },
+  linkedin: {
+    label: "LinkedIn",
+    order: 30,
+    category: "pipeline",
+    region: "global",
+  },
+  glassdoor: {
+    label: "Glassdoor",
+    order: 40,
+    category: "pipeline",
+    region: "global",
   },
-  indeed: { label: "Indeed", order: 20, category: "pipeline" },
-  linkedin: { label: "LinkedIn", order: 30, category: "pipeline" },
-  glassdoor: { label: "Glassdoor", order: 40, category: "pipeline" },
   ukvisajobs: {
     label: "UK Visa Jobs",
     order: 50,
     category: "pipeline",
     requiresCredentials: true,
     ukOnly: true,
+    region: "uk",
   },
   adzuna: {
     label: "Adzuna",
     order: 60,
     category: "pipeline",
     requiresCredentials: true,
+    region: "global",
   },
-  hiringcafe: { label: "Hiring Cafe", order: 70, category: "pipeline" },
+  hiringcafe: {
-  startupjobs: { label: "startup.jobs", order: 80, category: "pipeline" },
+    label: "Hiring Cafe",
-  manual: { label: "Manual", order: 90, category: "manual" },
+    order: 70,
+    category: "pipeline",
+    region: "global",
+  },
+  startupjobs: {
+    label: "startup.jobs",
+    order: 80,
+    category: "pipeline",
+    region: "global",
+  },
+  usajobs: {
+    label: "USAJOBS",
+    order: 110,
+    category: "pipeline",
+    requiresCredentials: true,
+    countryAllowlist: ["united states", "usa", "us"],
+    region: "us",
+  },
+  jobicy: {
+    label: "Jobicy (Remote)",
+    order: 120,
+    category: "pipeline",
+    region: "remote",
+  },
+  themuse: {
+    label: "The Muse",
+    order: 130,
+    category: "pipeline",
+    region: "global",
+  },
+  jooble: {
+    label: "Jooble",
+    order: 140,
+    category: "pipeline",
+    requiresCredentials: true,
+    region: "global",
+  },
+  careerjet: {
+    label: "Careerjet",
+    order: 150,
+    category: "pipeline",
+    requiresCredentials: true,
+    region: "global",
+  },
+  reed: {
+    label: "Reed",
+    order: 160,
+    category: "pipeline",
+    requiresCredentials: true,
+    ukOnly: true,
+    countryAllowlist: ["united kingdom", "uk", "great britain", "england"],
+    region: "uk",
+  },
+  remoteok: {
+    label: "Remote OK",
+    order: 170,
+    category: "pipeline",
+    region: "remote",
+  },
+  remotive: {
+    label: "Remotive",
+    order: 175,
+    category: "pipeline",
+    region: "remote",
+  },
+  arbeitnow: {
+    label: "Arbeitnow",
+    order: 180,
+    category: "pipeline",
+    region: "global",
+  },
+  himalayas: {
+    label: "Himalayas",
+    order: 185,
+    category: "pipeline",
+    region: "remote",
+  },
+  weworkremotely: {
+    label: "We Work Remotely",
+    order: 190,
+    category: "pipeline",
+    region: "remote",
+  },
+  fourdayweek: {
+    label: "4 Day Week",
+    order: 195,
+    category: "pipeline",
+    region: "remote",
+  },
+  ashby: {
+    label: "Ashby (ATS)",
+    order: 210,
+    category: "pipeline",
+    region: "global",
+  },
+  lever: {
+    label: "Lever (ATS)",
+    order: 220,
+    category: "pipeline",
+    region: "global",
+  },
+  greenhouse: {
+    label: "Greenhouse (ATS)",
+    order: 230,
+    category: "pipeline",
+    region: "global",
+  },
+  workday: {
+    label: "Workday (ATS)",
+    order: 240,
+    category: "pipeline",
+    region: "global",
+  },
+  manual: { label: "Manual", order: 900, category: "manual" },
 };
 
 export const PIPELINE_EXTRACTOR_SOURCE_IDS = EXTRACTOR_SOURCE_IDS.filter(

shared/src/index.ts

@@ -1,4 +1,5 @@
 export * from "./extractors";
+export * from "./job-fingerprint";
 export * from "./job-url-canonical";
 export * from "./location-support";
 export * from "./types";

shared/src/job-fingerprint.test.ts

@@ -0,0 +1,80 @@
+import { describe, expect, it } from "vitest";
+import {
+  buildJobContentFingerprint,
+  normalizeEmployerForFingerprint,
+  normalizeTitleForFingerprint,
+} from "./job-fingerprint";
+
+describe("buildJobContentFingerprint", () => {
+  it("collapses the same role across sources", () => {
+    const a = buildJobContentFingerprint({
+      employer: "Stripe, Inc.",
+      title: "Senior Software Engineer (Backend) - Toronto, ON",
+    });
+    const b = buildJobContentFingerprint({
+      employer: "stripe inc",
+      title: "Senior Software Engineer  (Backend)",
+    });
+    expect(a).toBe(b);
+    expect(a).not.toBeNull();
+  });
+
+  it("ignores trailing location decorations on titles", () => {
+    const a = buildJobContentFingerprint({
+      employer: "Acme",
+      title: "Software Engineer — Remote",
+    });
+    const b = buildJobContentFingerprint({
+      employer: "Acme",
+      title: "Software Engineer",
+    });
+    expect(a).toBe(b);
+  });
+
+  it("strips diacritics and punctuation", () => {
+    const a = buildJobContentFingerprint({
+      employer: "Café Münchën",
+      title: "Étudiant – Stage",
+    });
+    const b = buildJobContentFingerprint({
+      employer: "cafe munchen",
+      title: "Etudiant Stage",
+    });
+    expect(a).toBe(b);
+  });
+
+  it("returns null when employer or title is empty", () => {
+    expect(
+      buildJobContentFingerprint({ employer: "", title: "Engineer" }),
+    ).toBeNull();
+    expect(
+      buildJobContentFingerprint({ employer: "Acme", title: "" }),
+    ).toBeNull();
+  });
+
+  it("does not collapse different roles at the same employer", () => {
+    const a = buildJobContentFingerprint({
+      employer: "Acme",
+      title: "Software Engineer",
+    });
+    const b = buildJobContentFingerprint({
+      employer: "Acme",
+      title: "Product Designer",
+    });
+    expect(a).not.toBe(b);
+  });
+
+  describe("normalizers", () => {
+    it("normalizeEmployerForFingerprint strips legal suffixes", () => {
+      expect(normalizeEmployerForFingerprint("Acme Corporation")).toBe("acme");
+      expect(normalizeEmployerForFingerprint("Acme, LLC")).toBe("acme");
+      expect(normalizeEmployerForFingerprint("Acme GmbH")).toBe("acme");
+    });
+
+    it("normalizeTitleForFingerprint drops leading repost markers", () => {
+      expect(normalizeTitleForFingerprint("[Reposted] Software Engineer")).toBe(
+        "softwareengineer",
+      );
+    });
+  });
+});

shared/src/job-fingerprint.ts

@@ -0,0 +1,77 @@
+/**
+ * Cross-source duplicate detection.
+ *
+ * Two postings from different sources almost always describe the same role
+ * when their employer + title agree once you strip noise (case, punctuation,
+ * tracking suffixes, common decorations like "(Remote)", "- Toronto, ON" etc).
+ * The fingerprint is intentionally coarse so we err on the side of skipping
+ * a duplicate rather than re-showing it from a second source.
+ */
+
+const PUNCTUATION_RE = /[\p{P}\p{S}]+/gu;
+const WHITESPACE_RE = /\s+/g;
+const LEADING_NOISE_RE = /^(?:re-?post(?:ed)?|new|hot|urgent)\s*[-:]?\s*/i;
+const PARENS_RE = /\s*[([][^)\]]*[)\]]/g;
+// Trailing decorations we know are location / arrangement metadata, not role
+// suffix. Matched after the title body and stripped before fingerprinting.
+// Examples we want to strip:
+//   "Software Engineer — Remote"
+//   "Senior Engineer - Toronto, ON"
+//   "Designer | Hybrid"
+// Examples we must NOT strip (otherwise we'd collide unrelated roles):
+//   "Etudiant – Stage" (Stage is a role qualifier in French postings)
+//   "Designer — Senior" (level qualifier)
+const TRAILING_LOCATION_KEYWORDS_RE =
+  /\s+[-|–—]\s+(?:remote|hybrid|on[\s-]?site|wfh|telework|anywhere)\s*$/i;
+// Escape the ASCII hyphen explicitly so it doesn't form a character range
+// with the surrounding delimiters (which would silently swallow letters).
+const TRAILING_CITY_REGION_RE = /\s+[-|–—]\s+[^,\-|–—]+,\s*[^,\-|–—]+\s*$/;
+const COMPANY_LEGAL_SUFFIX_RE =
+  /\b(?:inc|inc\.|ltd|ltd\.|llc|gmbh|s\.a\.|s\.r\.l|sa|nv|bv|plc|corp|corporation|co|company|holdings|holding)\b/g;
+
+function stripDiacritics(input: string): string {
+  return input.normalize("NFKD").replace(/[\u0300-\u036f]/g, "");
+}
+
+export function normalizeEmployerForFingerprint(
+  employer: string | null | undefined,
+): string {
+  if (!employer) return "";
+  let value = stripDiacritics(employer.toLowerCase()).trim();
+  value = value.replace(PARENS_RE, " ");
+  value = value.replace(COMPANY_LEGAL_SUFFIX_RE, " ");
+  value = value.replace(PUNCTUATION_RE, " ");
+  value = value.replace(WHITESPACE_RE, "").trim();
+  return value;
+}
+
+export function normalizeTitleForFingerprint(
+  title: string | null | undefined,
+): string {
+  if (!title) return "";
+  let value = stripDiacritics(title.toLowerCase()).trim();
+  value = value.replace(LEADING_NOISE_RE, "");
+  value = value.replace(PARENS_RE, " ");
+  value = value.replace(TRAILING_LOCATION_KEYWORDS_RE, " ");
+  value = value.replace(TRAILING_CITY_REGION_RE, " ");
+  value = value.replace(PUNCTUATION_RE, " ");
+  value = value.replace(WHITESPACE_RE, "").trim();
+  return value;
+}
+
+/**
+ * Build a stable, source-agnostic fingerprint for a posting.
+ *
+ * Returns `null` when employer or title is empty after normalization, so
+ * callers fall back to URL/sourceJobId equality and don't accidentally
+ * collapse unrelated rows under the empty key.
+ */
+export function buildJobContentFingerprint(args: {
+  employer: string | null | undefined;
+  title: string | null | undefined;
+}): string | null {
+  const employer = normalizeEmployerForFingerprint(args.employer);
+  const title = normalizeTitleForFingerprint(args.title);
+  if (!employer || !title) return null;
+  return `${employer}::${title}`;
+}

shared/src/location-support.ts

@@ -99,7 +99,12 @@ export const SUPPORTED_COUNTRY_INPUTS = [
   "worldwide",
 ] as const;
 
-const UK_ONLY_SOURCES = new Set<JobSource>(["gradcracker", "ukvisajobs"]);
+const UK_ONLY_SOURCES = new Set<JobSource>([
+  "gradcracker",
+  "ukvisajobs",
+  "reed",
+]);
+const US_ONLY_SOURCES = new Set<JobSource>(["usajobs"]);
 const GLASSDOOR_SUPPORTED_COUNTRIES = new Set(
   [
     "australia",
@@ -170,6 +175,10 @@ export function isUkCountry(country: string | null | undefined): boolean {
   return normalizeCountryKey(country) === "united kingdom";
 }
 
+export function isUsCountry(country: string | null | undefined): boolean {
+  return normalizeCountryKey(country) === "united states";
+}
+
 export function isGlassdoorCountry(
   country: string | null | undefined,
 ): boolean {
@@ -187,6 +196,7 @@ export function isSourceAllowedForCountry(
   country: string | null | undefined,
 ): boolean {
   if (UK_ONLY_SOURCES.has(source)) return isUkCountry(country);
+  if (US_ONLY_SOURCES.has(source)) return isUsCountry(country);
   if (source === "glassdoor") return isGlassdoorCountry(country);
   if (source === "adzuna") return getAdzunaCountryCode(country) !== null;
   return true;

shared/src/settings-registry.ts

@@ -28,6 +28,24 @@ function parseJsonArrayOrNull(raw: string | undefined): string[] | null {
   }
 }
 
+/**
+ * Parse a delimited list (comma / newline / pipe) into a deduped, trimmed
+ * array. Used for env-backed defaults like LEVER_COMPANIES="acme,stripe".
+ */
+function parseCompanyList(raw: string | undefined | null): string[] {
+  if (!raw) return [];
+  const out: string[] = [];
+  const seen = new Set<string>();
+  for (const piece of raw.split(/[\n,;|]+/)) {
+    const value = piece.trim();
+    if (!value) continue;
+    if (seen.has(value)) continue;
+    seen.add(value);
+    out.push(value);
+  }
+  return out;
+}
+
 function parseBitBoolOrNull(raw: string | undefined): boolean | null {
   if (!raw) return null;
   return raw === "true" || raw === "1";
@@ -336,6 +354,145 @@ export const settingsRegistry = {
     parse: parseIntOrNull,
     serialize: serializeNullableNumber,
   },
+  // --- New extractor caps & per-source target lists ---
+  usajobsMaxJobsPerTerm: {
+    kind: "typed" as const,
+    schema: z.number().int().min(1).max(1000),
+    default: (): number =>
+      parseInt(
+        typeof process !== "undefined"
+          ? process.env.USAJOBS_MAX_JOBS_PER_TERM || "100"
+          : "100",
+        10,
+      ),
+    parse: parseIntOrNull,
+    serialize: serializeNullableNumber,
+  },
+  jobicyMaxJobsPerTerm: {
+    kind: "typed" as const,
+    schema: z.number().int().min(1).max(1000),
+    default: (): number => 100,
+    parse: parseIntOrNull,
+    serialize: serializeNullableNumber,
+  },
+  themuseMaxJobsPerTerm: {
+    kind: "typed" as const,
+    schema: z.number().int().min(1).max(1000),
+    default: (): number => 100,
+    parse: parseIntOrNull,
+    serialize: serializeNullableNumber,
+  },
+  joobleMaxJobsPerTerm: {
+    kind: "typed" as const,
+    schema: z.number().int().min(1).max(1000),
+    default: (): number => 100,
+    parse: parseIntOrNull,
+    serialize: serializeNullableNumber,
+  },
+  careerjetMaxJobsPerTerm: {
+    kind: "typed" as const,
+    schema: z.number().int().min(1).max(1000),
+    default: (): number => 100,
+    parse: parseIntOrNull,
+    serialize: serializeNullableNumber,
+  },
+  reedMaxJobsPerTerm: {
+    kind: "typed" as const,
+    schema: z.number().int().min(1).max(1000),
+    default: (): number => 100,
+    parse: parseIntOrNull,
+    serialize: serializeNullableNumber,
+  },
+  remoteokMaxJobsPerTerm: {
+    kind: "typed" as const,
+    schema: z.number().int().min(1).max(1000),
+    default: (): number => 100,
+    parse: parseIntOrNull,
+    serialize: serializeNullableNumber,
+  },
+  remotiveMaxJobsPerTerm: {
+    kind: "typed" as const,
+    schema: z.number().int().min(1).max(1000),
+    default: (): number => 100,
+    parse: parseIntOrNull,
+    serialize: serializeNullableNumber,
+  },
+  arbeitnowMaxJobsPerTerm: {
+    kind: "typed" as const,
+    schema: z.number().int().min(1).max(1000),
+    default: (): number => 100,
+    parse: parseIntOrNull,
+    serialize: serializeNullableNumber,
+  },
+  himalayasMaxJobsPerTerm: {
+    kind: "typed" as const,
+    schema: z.number().int().min(1).max(1000),
+    default: (): number => 100,
+    parse: parseIntOrNull,
+    serialize: serializeNullableNumber,
+  },
+  weworkremotelyMaxJobsPerTerm: {
+    kind: "typed" as const,
+    schema: z.number().int().min(1).max(1000),
+    default: (): number => 100,
+    parse: parseIntOrNull,
+    serialize: serializeNullableNumber,
+  },
+  fourdayweekMaxJobsPerTerm: {
+    kind: "typed" as const,
+    schema: z.number().int().min(1).max(1000),
+    default: (): number => 100,
+    parse: parseIntOrNull,
+    serialize: serializeNullableNumber,
+  },
+  /**
+   * Comma- or newline-separated company slugs to fetch from public ATS feeds.
+   * `lever`, `ashby`, and `greenhouse` each take one entry per company.
+   */
+  leverCompanies: {
+    kind: "typed" as const,
+    schema: z.array(z.string().trim().min(1).max(100)).max(200),
+    default: (): string[] =>
+      parseCompanyList(
+        typeof process !== "undefined" ? process.env.LEVER_COMPANIES : "",
+      ),
+    parse: parseJsonArrayOrNull,
+    serialize: serializeNullableJsonArray,
+  },
+  ashbyCompanies: {
+    kind: "typed" as const,
+    schema: z.array(z.string().trim().min(1).max(100)).max(200),
+    default: (): string[] =>
+      parseCompanyList(
+        typeof process !== "undefined" ? process.env.ASHBY_COMPANIES : "",
+      ),
+    parse: parseJsonArrayOrNull,
+    serialize: serializeNullableJsonArray,
+  },
+  greenhouseCompanies: {
+    kind: "typed" as const,
+    schema: z.array(z.string().trim().min(1).max(100)).max(200),
+    default: (): string[] =>
+      parseCompanyList(
+        typeof process !== "undefined" ? process.env.GREENHOUSE_COMPANIES : "",
+      ),
+    parse: parseJsonArrayOrNull,
+    serialize: serializeNullableJsonArray,
+  },
+  /**
+   * Workday tenant configurations as JSON, e.g.
+   * `[{"company":"Acme","tenantUrl":"https://acme.wd1.myworkdayjobs.com","sites":["External"]}]`.
+   */
+  workdayTenants: {
+    kind: "typed" as const,
+    schema: z.array(z.string().trim().min(1).max(2000)).max(50),
+    default: (): string[] =>
+      parseCompanyList(
+        typeof process !== "undefined" ? process.env.WORKDAY_TENANTS : "",
+      ),
+    parse: parseJsonArrayOrNull,
+    serialize: serializeNullableJsonArray,
+  },
   searchTerms: {
     kind: "typed" as const,
     schema: z.array(z.string().trim().min(1).max(200)).max(100),
@@ -626,6 +783,40 @@ export const settingsRegistry = {
     envKey: "ADZUNA_APP_ID",
     schema: z.string().trim().max(200),
   },
+  // --- New extractor keys / identifiers (non-secret) ---
+  usajobsUserAgent: {
+    kind: "string" as const,
+    envKey: "USAJOBS_USER_AGENT",
+    schema: z.string().trim().max(200),
+  },
+  themuseApiKey: {
+    kind: "string" as const,
+    envKey: "THEMUSE_API_KEY",
+    schema: z.string().trim().max(200),
+  },
+  /** Publisher API key (Basic auth user); Careerjet labels this “API key” in the dashboard. */
+  careerjetAffid: {
+    kind: "string" as const,
+    envKey: "CAREERJET_AFFID",
+    schema: z.string().trim().max(200),
+  },
+  /** Required Referer URL for v4 (your job-search page that triggers API use). */
+  careerjetReferer: {
+    kind: "string" as const,
+    envKey: "CAREERJET_REFERER",
+    schema: z.string().trim().max(500),
+  },
+  /** Must match an IP allowlisted in Careerjet (usually your server egress IP). */
+  careerjetUserIp: {
+    kind: "string" as const,
+    envKey: "CAREERJET_USER_IP",
+    schema: z.string().trim().max(80),
+  },
+  careerjetUserAgent: {
+    kind: "string" as const,
+    envKey: "CAREERJET_USER_AGENT",
+    schema: z.string().trim().max(500),
+  },
   basicAuthUser: {
     kind: "string" as const,
     envKey: "BASIC_AUTH_USER",
@@ -658,6 +849,22 @@ export const settingsRegistry = {
     envKey: "ADZUNA_APP_KEY",
     schema: z.string().trim().max(2000),
   },
+  // --- Secrets for new extractors ---
+  usajobsApiKey: {
+    kind: "secret" as const,
+    envKey: "USAJOBS_API_KEY",
+    schema: z.string().trim().max(2000),
+  },
+  joobleApiKey: {
+    kind: "secret" as const,
+    envKey: "JOOBLE_API_KEY",
+    schema: z.string().trim().max(2000),
+  },
+  reedApiKey: {
+    kind: "secret" as const,
+    envKey: "REED_API_KEY",
+    schema: z.string().trim().max(2000),
+  },
   basicAuthPassword: {
     kind: "secret" as const,
     envKey: "BASIC_AUTH_PASSWORD",

shared/src/testing/factories.ts

@@ -188,6 +188,22 @@ export const createAppSettings = (
   adzunaMaxJobsPerTerm: { value: 50, default: 50, override: null },
   gradcrackerMaxJobsPerTerm: { value: 50, default: 50, override: null },
   startupjobsMaxJobsPerTerm: { value: 50, default: 50, override: null },
+  usajobsMaxJobsPerTerm: { value: 50, default: 50, override: null },
+  jobicyMaxJobsPerTerm: { value: 50, default: 50, override: null },
+  themuseMaxJobsPerTerm: { value: 50, default: 50, override: null },
+  joobleMaxJobsPerTerm: { value: 50, default: 50, override: null },
+  careerjetMaxJobsPerTerm: { value: 50, default: 50, override: null },
+  reedMaxJobsPerTerm: { value: 50, default: 50, override: null },
+  remoteokMaxJobsPerTerm: { value: 50, default: 50, override: null },
+  remotiveMaxJobsPerTerm: { value: 50, default: 50, override: null },
+  arbeitnowMaxJobsPerTerm: { value: 50, default: 50, override: null },
+  himalayasMaxJobsPerTerm: { value: 50, default: 50, override: null },
+  weworkremotelyMaxJobsPerTerm: { value: 50, default: 50, override: null },
+  fourdayweekMaxJobsPerTerm: { value: 50, default: 50, override: null },
+  leverCompanies: { value: [], default: [], override: null },
+  ashbyCompanies: { value: [], default: [], override: null },
+  greenhouseCompanies: { value: [], default: [], override: null },
+  workdayTenants: { value: [], default: [], override: null },
   searchTerms: {
     value: ["Software Engineer"],
     default: ["Software Engineer"],
@@ -256,6 +272,15 @@ export const createAppSettings = (
   adzunaAppId: null,
   adzunaAppKeyHint: null,
   webhookSecretHint: null,
+  usajobsUserAgent: null,
+  themuseApiKey: null,
+  careerjetAffid: null,
+  careerjetReferer: null,
+  careerjetUserIp: null,
+  careerjetUserAgent: null,
+  usajobsApiKeyHint: null,
+  joobleApiKeyHint: null,
+  reedApiKeyHint: null,
   basicAuthActive: false,
   localResumeFileConfigured: false,
   backupEnabled: { value: false, default: false, override: null },

shared/src/types/settings.ts

@@ -201,6 +201,22 @@ export interface AppSettings {
   adzunaMaxJobsPerTerm: Resolved<number>;
   gradcrackerMaxJobsPerTerm: Resolved<number>;
   startupjobsMaxJobsPerTerm: Resolved<number>;
+  usajobsMaxJobsPerTerm: Resolved<number>;
+  jobicyMaxJobsPerTerm: Resolved<number>;
+  themuseMaxJobsPerTerm: Resolved<number>;
+  joobleMaxJobsPerTerm: Resolved<number>;
+  careerjetMaxJobsPerTerm: Resolved<number>;
+  reedMaxJobsPerTerm: Resolved<number>;
+  remoteokMaxJobsPerTerm: Resolved<number>;
+  remotiveMaxJobsPerTerm: Resolved<number>;
+  arbeitnowMaxJobsPerTerm: Resolved<number>;
+  himalayasMaxJobsPerTerm: Resolved<number>;
+  weworkremotelyMaxJobsPerTerm: Resolved<number>;
+  fourdayweekMaxJobsPerTerm: Resolved<number>;
+  leverCompanies: Resolved<string[]>;
+  ashbyCompanies: Resolved<string[]>;
+  greenhouseCompanies: Resolved<string[]>;
+  workdayTenants: Resolved<string[]>;
   searchTerms: Resolved<string[]>;
   workplaceTypes: Resolved<Array<"remote" | "hybrid" | "onsite">>;
   blockedCompanyKeywords: Resolved<string[]>;
@@ -241,6 +257,12 @@ export interface AppSettings {
   ukvisajobsEmail: string | null;
   adzunaAppId: string | null;
   basicAuthUser: string | null;
+  usajobsUserAgent: string | null;
+  themuseApiKey: string | null;
+  careerjetAffid: string | null;
+  careerjetReferer: string | null;
+  careerjetUserIp: string | null;
+  careerjetUserAgent: string | null;
 
   // Secret hints:
   llmApiKeyHint: string | null;
@@ -250,6 +272,9 @@ export interface AppSettings {
   adzunaAppKeyHint: string | null;
   basicAuthPasswordHint: string | null;
   webhookSecretHint: string | null;
+  usajobsApiKeyHint: string | null;
+  joobleApiKeyHint: string | null;
+  reedApiKeyHint: string | null;
 
   // Computed:
   basicAuthActive: boolean;