Probity Vectors
Test an agent-evidence verifier against saved records with known outcomes. The corpora cover valid records, altered bytes, missing evidence and open specification interpretations.
The recording is demo/four-arms.sh from a fresh clone; vhs demo/four-arms.tape re-records it.
Try it
From a reviewed source checkout, build the Go verifier and a wheel, then run the installed harness outside the checkout. You need Go 1.26+, Python 3.11+ and uv. Record the checkout's full commit ID with your report.
git rev-parse HEAD
run_dir="$(mktemp -d)"
GOWORK=off go build -o "$run_dir/aee-verify" ./cmd/aee-verify
uv build --wheel --out-dir "$run_dir"
wheel="$(find "$run_dir" -maxdepth 1 -name 'agent_evidence_vectors-*.whl')"
cd "$run_dir"
uv run --no-project --with "$wheel" agent-evidence-vectors --verifier "$run_dir/aee-verify --json"
Open conformance-report.json for the per-vector results. A complete external run has rail: external, verifier.vectorsExecuted equal to totals.vectors, and zero totals.suiteRefusals. The report separates conformance from reason-code agreement.
See installation routes and release identity for package availability and signature checks.
Select v0.18.1 on GitHub and PyPI. Install the matching release with Go's executable directory on your PATH:
go install github.com/probityai/agent-evidence-vectors/cmd/aee-verify@v0.18.1
uvx agent-evidence-vectors==0.18.1 --verifier "aee-verify --json"
Test your verifier
The harness calls your executable with a vector-file path. Return a verdict through the exit status and a JSON object on the last line of stdout. The verifier contract covers output fields, key policy and comparison rules.
Use the GitHub Action to run the same check in your build and retain its report. The CI guide gives a workflow step and the available inputs.
Verifiers that pin this corpus
agmi, Rul1an/aee-checker and MoltyCel/moltrust-api run a corpus from this repository in their own CI at a pinned release or commit, and this repository's CI runs agmi's anchored-chain verifier back against vectors-anchored-chain. The verifier table gives each one's corpus, pin and CI run.
Choose your next step
| I want to | Read |
|---|---|
| Pick a format and corpus | Corpus guide |
| Implement the verifier interface | Verifier contract |
| Understand what the fixtures test | Corpus measurements |
| Check release bytes and signatures | Release verification |
| Report a run or disagreement | Run reporting |
| See who runs and uses the corpora | Independent runs, Adopters |
Each corpus manifest records its source revision, digest and expected outcomes. DISTRIBUTION.md lists the shipped corpora and installation routes.
For fixture coverage, see Corpus measurements. To check signed release bytes, use Release verification.
Source maps: repository layout, corpus readers and witness attestor.
Contribute
Open an issue with the corpus revision, verifier build and report. For patches, start with CONTRIBUTING.md. For citations, use the citation guide.
Need to check a supplied claim against saved evidence? See Probity Verify. Browse the Probity projects for records, vocabulary and admission tools.
Apache-2.0. LICENSE.
Metadata
Release files for agent-evidence-vectors 0.19.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | |
|---|---|---|---|---|
| agent_evidence_vectors-0.19.0.tar.gz | 1.9 MB | With provenance | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| agent_evidence_vectors-0.19.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 5.9 MB
Release files / agent_evidence_vectors-0.19.0.tar.gz
| Download URL | agent_evidence_vectors-0.19.0.tar.gz |
|---|---|
| Size | 1.9 MB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
092fc2020fce0d484b0a8a4bf62274b69c7faf36d30a479d746814ae8a41ebd3
|
|
BLAKE2b-256 checksum How to use checksums |
ce5c05f746a62b698277e0b67f1c2985739629a196add08be7e6544cf4f538a0
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 17, 2026.
Transparency logRelease files / agent_evidence_vectors-0.19.0-py3-none-any.whl
| Download URL | agent_evidence_vectors-0.19.0-py3-none-any.whl |
|---|---|
| Size | 4.0 MB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
244983c959490f54ab2bfd09f30e7b98ac69ab9def770c264b4ef7b0fc911d48
|
|
BLAKE2b-256 checksum How to use checksums |
1049258a703bd178d7d3d3a4ec117e274667c50bc30cb5235a9dc75f32926e6b
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 17, 2026.
Transparency logSecurity for agent-evidence-vectors 0.19.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
Full provenance
All 2 files in this release have been uploaded with provenance metadata.
Attestation for agent_evidence_vectors-0.19.0.tar.gz
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 17, 2026.
Transparency logA note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.