The safeguard library
See the safeguards before you use them.
A Control Pack is a reusable set of safeguard rules. Each one explains when it applies, which decisions need you, what to implement, and what evidence to record. Some also include implementation recipes and pre-release scenarios.
Want to see how a pack behaves? Open the reproducible verification directory to inspect and run one published example for every Control Pack.
We could not load the safeguard library. Please try again.
How these safeguards work
Control Packs provide structured design requirements, implementation recipes, pre-release scenarios, and evidence expectations for AI-assisted development. The same versioned data appears in the review, API, MCP tools, and downloaded bundle. Related guidance links are contextual only; packs in review are published for transparency and feedback, not as certification or independent verification.