# ============================================================================
# LLM Council Configuration
# ============================================================================
# Copy this file to .env and customize as needed
# ============================================================================

# ----------------------------------------------------------------------------
# Provider & Server Configuration
# ----------------------------------------------------------------------------

# Use local Ollama (automatically sets base URL to http://localhost:11434)
USE_LOCAL_OLLAMA=true

# For remote Ollama or other OpenAI-compatible servers, comment USE_LOCAL_OLLAMA
# and set the base URL instead:
# OPENAI_COMPAT_BASE_URL=http://your-server:11434
# OPENAI_COMPAT_BASE_URL=http://10.0.30.63:8000

# Optional API key if your server requires authentication
# OPENAI_COMPAT_API_KEY=your_api_key_here

# ----------------------------------------------------------------------------
# Model Configuration
# ----------------------------------------------------------------------------

# Models for the council (comma or newline separated)
# Tip: Check available models with: curl -s 'http://localhost:8001/api/llm/status?probe=true'
COUNCIL_MODELS=llama3.2:3b,qwen2.5:3b,gemma2:2b

# Chairman model (synthesizes final response from council)
CHAIRMAN_MODEL=llama3.2:3b

# Maximum tokens per request
MAX_TOKENS=1024

# ----------------------------------------------------------------------------
# Timeout Configuration
# ----------------------------------------------------------------------------

# Default timeout for general LLM queries (Stage 1: council responses)
LLM_TIMEOUT_SECONDS=300.0

# Timeout for chairman synthesis (may need longer for complex responses)
CHAIRMAN_TIMEOUT_SECONDS=180.0

# Timeout for title generation (short responses)
TITLE_GENERATION_TIMEOUT_SECONDS=120.0

# HTTP client timeout for OpenAI-compatible server (fallback, rarely used)
OPENAI_COMPAT_TIMEOUT_SECONDS=300

# ----------------------------------------------------------------------------
# Retry Configuration
# ----------------------------------------------------------------------------

# Number of retries for failed requests (retryable HTTP errors)
OPENAI_COMPAT_RETRIES=2

# Exponential backoff base delay between retries (seconds)
OPENAI_COMPAT_RETRY_BACKOFF_SECONDS=0.5

# ----------------------------------------------------------------------------
# Concurrency Configuration
# ----------------------------------------------------------------------------

# Maximum concurrent model requests (0 = unlimited, 1 = sequential)
LLM_MAX_CONCURRENCY=1

# ----------------------------------------------------------------------------
# Document Upload Configuration (Optional)
# ----------------------------------------------------------------------------

# Directory for uploaded markdown documents (per-conversation)
# DOCS_DIR=data/docs

# Maximum document size in bytes (default: 1MB)
# MAX_DOC_BYTES=1000000

# Maximum characters to preview when fetching documents (default: 20000)
# MAX_DOC_PREVIEW_CHARS=20000