9 projects
dprovenancekit
Regression testing for AI agents — fail the PR when your agent skips a step, even when the answer still looks right.
warrantkit
Runtime enforcement and proof platform for autonomous AI agents
agentcontainment
Controller-owned authorization and independently verifiable containment for AI agents
claimproofkit
Apache-2.0 licensed, evidence-bound claim verification that blocks generated documents whose material claims the supplied evidence does not support. Python implementation of Claim Proof Specification v1, in locked parity with the Swift reference implementation.
cancelscope
Cooperative cancellation, timeouts, and cancel scopes for synchronous Python
pytest-flakedoctor
The flaky-test doctor: diagnoses WHY a test is flaky and hands you a deterministic reproduction
interleave-test
Deterministic thread-interleaving fuzzer: loom-style concurrency testing for free-threaded Python
hermetic-sandbox
One context manager that freezes time, seeds all randomness, isolates the filesystem, and blocks network — keyed to a single replayable seed, with record/replay of nondeterminism from failing runs
DProvenanceKitPython
Compatibility shim that installs dprovenancekit.