Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
76 lines (62 loc) · 3.34 KB

CONTRIBUTING.md

File metadata and controls

76 lines (62 loc) · 3.34 KB
Raw

Contributing to LightAgent

Thanks for helping improve LightAgent. The project values small, focused changes that preserve the lightweight core and keep existing APIs compatible.

Development Setup

python -m venv .venv
source .venv/bin/activate
python -m pip install --upgrade pip
python -m pip install -r requirements.txt
python -m pip install pytest

Before Opening a Pull Request

Run the focused regression suite:

python -m compileall -q LightAgent
PYTHONPATH=. python -m pytest -q tests/test_v065_core.py tests/test_v070_tracing.py tests/test_memory_policy.py

Contribution Guidelines

  • Keep default agent.run() behavior backward compatible unless a breaking change is explicitly approved and documented.
  • Prefer focused modules and tests over adding more logic to LightAgent/core.py.
  • Do not include credentials, local logs, generated build artifacts, or private tool implementations.
  • For memory, MCP, tool loading, or code execution changes, include a short security note in the PR.
  • Update README or docs/ when behavior, configuration, or public APIs change.

Third-Party Services And Provider Documentation

Contributions that add or replace a hosted model provider, gateway, external API, badge, chart, or other remote service must make the dependency and its trust boundary reviewable.

  • Disclose any employment, ownership, sponsorship, or other affiliation with the service in the issue and pull request.
  • Link the service's official, publicly accessible API documentation, model or feature catalog, and source repository when the service is open source.
  • Prefer provider-neutral OpenAI-compatible configuration guidance. A dedicated provider section should document a meaningful compatibility detail that is not already covered by model, api_key, and base_url.
  • Use stable example identifiers where possible. When model names or features are controlled by the service, link the live catalog and state that availability may change.
  • Read credentials from environment variables. Never include working keys, encrypted tokens, account identifiers, or credentials embedded in public URLs.
  • Include reproducible compatibility evidence for the behavior being claimed. For model gateways, cover chat completions and any claimed streaming or tool calling support. Clearly separate local or CI validation from credentialed live-service validation.
  • Limit project documentation to verifiable configuration and compatibility facts. Do not add unverified pricing, model-count, availability, performance, failover, security, privacy, or regulatory guarantees.
  • For externally hosted images or embeds, identify the operator and source, describe what request data leaves GitHub, and explain why a repository-owned static asset is not sufficient.
  • External-contributor workflows may remain unapproved until maintainers finish reviewing the proposed service dependency. CI approval does not imply project endorsement of that service.

Maintainers may decline or remove a dedicated listing when its claims cannot be independently verified, its maintenance burden is disproportionate, or generic OpenAI-compatible documentation already covers the integration.

Reporting Issues

Please use the bug or feature templates when possible. Security issues should be reported privately through GitHub Security Advisories.