Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
60 lines (42 loc) · 1.71 KB

CONTRIBUTING.md

File metadata and controls

60 lines (42 loc) · 1.71 KB

Contributing

Thanks for helping improve agent-memory.

Development setup

git clone https://github.com/cafitac/agent-memory.git
cd agent-memory
uv run agent-memory --help
uv run pytest tests/ -q

The project uses:

  • Python 3.11+
  • uv for local development commands
  • npm only for the thin launcher package and package smoke checks
  • SQLite for local storage
  • pytest for tests

Before opening a PR

Run:

uv run pytest tests/ -q
uv run python scripts/check_release_metadata.py
uv run python scripts/smoke_release_readiness.py
npm pack --dry-run

If your change affects install, release, or launcher behavior, also run a clean install smoke from outside the source checkout. See docs/install-smoke.md.

Change expectations

  • Keep user-facing behavior documented in README.md.
  • Keep security/privacy-impacting changes reflected in SECURITY.md or PRIVACY.md.
  • Add or update tests for behavior changes.
  • Keep retrieval evaluator changes covered in tests/test_retrieval_evaluation.py.
  • Prefer small PRs with one clear purpose.
  • Do not commit local memory databases, virtualenvs, agent state directories, or secrets.

Retrieval evaluation changes

When changing retrieval behavior or retrieval-eval report fields:

uv run pytest tests/test_retrieval_evaluation.py -q
uv run pytest tests/test_retrieval_trace.py -q
uv run pytest tests/ -q

The evaluator should continue to exercise the real retrieval path, preserve JSON as the machine-readable contract, and keep text reports human-actionable.

Reporting issues

For ordinary bugs, use the bug report issue template.

For vulnerabilities or private data exposure, follow SECURITY.md instead of posting exploit details publicly.