Skip to content

configuration.md

Latest commit

 

History

History
29 lines (22 loc) · 1.08 KB

configuration.md

File metadata and controls

29 lines (22 loc) · 1.08 KB

Configuration

All configuration is via environment variables or a .env file in the project root.

Environment variables

Variable Required Default Description
OPENAI_API_KEY Yes — LLM API key
OPENAI_BASE_URL No https://api.deepseek.com/v1 LLM endpoint base URL
MODEL No deepseek-v4-pro Model name
CONCURRENCY No 30 Max concurrent LLM calls
TEMP_DIR No /tmp/standalone-trace-analyzer Temp storage for run outputs

Example .env

OPENAI_API_KEY=sk-...
OPENAI_BASE_URL=https://api.deepseek.com/v1
MODEL=deepseek-v4-pro
CONCURRENCY=30
TEMP_DIR=/tmp/standalone-trace-analyzer

Notes

  • OPENAI_BASE_URL and OPENAI_API_KEY use OpenAI-compatible naming for compatibility with the client library, but any OpenAI-compatible provider works.
  • TEMP_DIR must be writable. Stage outputs are saved as JSON files under {TEMP_DIR}/runs/{run_id}/.
  • CONCURRENCY controls the asyncio.Semaphore limit on concurrent LLM requests. Lower this if you hit rate limits.