# Contributing

Thank you for helping keep this list useful.

## What belongs here

An entry belongs if it helps make the actions of an AI agent system traceable, auditable or reproducible: tracing and observability, durable or replayable execution, policy enforcement with an audit trail, tamper-evident logs and provenance, standards for any of these, or research that studies them.

General agent frameworks, model APIs and evaluation leaderboards belong only when traceability is a first-class part of what they do.

## Entry rules

- One entry per pull request, with a short note on what it makes traceable.
- Format: `- [Name](link) - Description.` One line, one or two sentences, ending with a period.
- Link the canonical repository, specification or paper page.
- Describe what the project does in neutral words. No marketing terms, no performance figures without a source, no tracking parameters.
- The project must be public and documented. Open source is preferred; hosted products are accepted when their documentation backs the description.
- Add the entry to the most fitting section, in alphabetical order. Propose a new section in the pull request if none fits.

## Affiliation

Self-submissions are welcome. Say in the pull request if you build, maintain or work for the project. The maintainers' own projects are kept in their own section, "From the maintainers", and never in the sections above it.

## Removal

Open an issue or a pull request if an entry is archived, moved, or no longer matches its description, with the evidence.

## License

By contributing, you agree that your contribution is released under [CC0 1.0](LICENSE).