# pistra

> pistra is the AI enforcement layer. Every model, tool and agent call is checked against your policy in line. Every decision leaves a signed record you can prove to anyone.

Written for an operator putting pistra in front of real traffic. `llms-full.txt` is everything; `llms-small.txt` drops the generated reference pages, which are long and are better fetched on demand from the set below.

pistra is source-available under BUSL-1.1 and self-hosted. There is no account to create and no hosted control plane.

## Documentation Sets

- [Abridged documentation](https://pistra.ai/llms-small.txt): a compact version of the documentation for pistra, with non-essential content removed
- [Complete documentation](https://pistra.ai/llms-full.txt): the full documentation for pistra
- [Reference](https://pistra.ai/_llms-txt/reference.txt): every configuration key, CLI command, policy name and detected entity type, generated from the code
- [How it works](https://pistra.ai/_llms-txt/how-it-works.txt): the design decisions, their limits, and what is deliberately not claimed
- [Guides](https://pistra.ai/_llms-txt/guides.txt): task-shaped instructions for a running deployment

## Notes

- The complete documentation includes all content from the official documentation
- The content is automatically generated from the same source as the official documentation

## Optional

- [pistra.ai](https://pistra.ai/): what it is, in one page
- [Product](https://pistra.ai/product/): the request path, what goes wrong on it, and the threat model for each of those things
- [Compared with other AI gateways and governance platforms](https://pistra.ai/compare/): where pistra sits beside the tools it is usually weighed against
- [Guard models measured in ten languages](https://pistra.ai/research/): what the shipped prompt-injection classifiers do to Arabic and eight other languages, and which one to pin