faultgate 0.3.1 · fixes from the first outside test, 180 real tau2-bench runsRead the changelog →
FAULTGATE DOCS · v0.3.1 · MIT · PYTHON ≥ 3.11

faultgate documentation

faultgate is a local-first release gate for AI agents. Hand it the traces your agent recorded, it checks every run for known failure modes, measures your run-to-run noise, and exits non-zero only when a candidate is a real regression.

START HERE

QuickstartInstall, run the built-in checks over your traces, and read the exit code.Business rulesTool allowlists, call budgets, argument limits, caller identity and prerequisites in policy.json.Gate & statisticsBuild a noise band from baseline runs and get PASS, FAIL or INCONCLUSIVE with the sample size.CI & GitHub ActionGate pull requests, post one PR comment, and upload the faultgate-audit artifact.

ALL PAGES

Getting Started

Evaluations & Rules

Data & Privacy

Integration & Reference