Skip to main content

agent-evidence-verify

PyPI

An offline verifier that decides whether a claim holds against evidence bytes the consumer pinned, with adapters for source passages, operand lineage, authority captures, bounded absence and spend journals.

It's for auditors, incident reviewers and relying parties who receive a claim about an AI system and need a decision they can replay, not a report they have to trust.

Install the released package: pip install agent-evidence-verify==0.5.0.

Quick start

Check a sample claim with the released package. You need Python 3.12 or newer and uv; nothing to clone. Version 0.5.0 ships the sample case and policies inside the wheel:

samples="$(uvx --from agent-evidence-verify==0.5.0 python -c 'import importlib.resources as r; print(r.files("agent_evidence_verify") / "samples")')"
uvx agent-evidence-verify==0.5.0 "$samples/case.json" --policy "$samples/policy-june.json"

It prints a decision packet that opens:

Probity decision packet v1
Case: "source-coverage-demo"
Claim: source_text_coverage/v1
Decision: contradicted
Reason: selected_span_missing_from_record
Policy SHA-256: dfda1551714f92702169d1411af6b0bfeee54732f934ce71265534af775fa388

A selected passage is in the source capture but missing from the stored report. The packet goes on to list every artifact binding and check. Swap in policy-april.json for not_established, or case-complete.json with policy-complete.json for supported. To run from a pinned source checkout instead, see Install a pinned version.

Use the decision in a check

All valid decisions exit 0, including contradicted and not_established. Exit 2 reports invalid input without a verdict. Require support for this claim by reading decision:

uvx agent-evidence-verify==0.5.0 "$samples/case.json" \
  --policy "$samples/policy-june.json" --json > decision.json &&
python3 - <<'PY'
import json

with open("decision.json") as stream:
    result = json.load(stream)
print(result["decision"], result["reason"])
raise SystemExit(0 if result["decision"] == "supported" else 1)
PY

This guard prints contradicted selected_span_missing_from_record and exits 1. supported means the stated claim passes the pinned policy's checks. Your application still owns authorization.

Status

Version 0.5.0 is on PyPI and ships the sample inputs used above. See release instructions for checked artifacts and the changelog for what each release changed. Fourteen adapters ship in 0.5.0: brokered_file_write/v1, source_text_coverage/v1, operand_lineage/v1, authority_anchor/v1, event_absence/v1, event_absence/v2, spend_reservation/v1, the done-receipt checks for coding agents protected_paths/v1, acceptance_observed/v1, tool_claim_match/v1, completion_claim/v1 and probe_observed/v1, adversarial_execution_evidence/v1, and done_receipt/v2, which derives a done receipt's two verdicts again from them: whether the work is done and whether the agent gamed the process. Each decision is supported, contradicted or not_established. The verifier checks supplied bytes only: it does not fetch URLs or authenticate the authority a policy names.

Documentation

page read it for
Start a project a template that checks an observer-recorded write with this verifier, pinned and in CI
Install a pinned version install outside the source checkout and replay three sample decisions
Choose a claim check select a task from checked recipes and retrieve exact commands and input digests
Select a consumer adapter supply exact adapter versions through adjudication, bundle creation and replay
Check a customer offer compare a draft with raw company records and an exact manager approval
Running the verifier every example command, the case and policy inputs, and what each decision means
Replayable decisions and the index ZIP bundles, replay, and the local index of decisions, challenges and supersessions
Adapters what operand lineage, authority anchor and event absence check, and what they leave to the consumer
Done-receipt checks check a coding agent's "done" against protected paths, an outside acceptance grade and its claimed command runs, with vantage and witness scope on every verdict
Spend reservations replay a pinned budget journal, refuse dispatch without an affordable reservation, and keep pending calls charged
Observation vantage select the producer and observation scope, replay bounded absence, and operate a witness with separate keys and retained heads
Adapter plan the architecture and the adapters planned next
A2A server and client send raw case bytes to an operator-configured verifier, or read a pinned Lab record
Source coverage, operand-lineage and done-check vectors conformance cases with pinned bytes and expected decisions

License

Apache-2.0.

Metadata

Release files for agent-evidence-verify 0.5.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for agent-evidence-verify 0.5.1
File Size Uploaded Provenance Actions
agent_evidence_verify-0.5.1.tar.gz 968.8 kB Without provenance Details for agent_evidence_verify-0.5.1.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for agent-evidence-verify 0.5.1
File Interpreter ABI Platform Actions
agent_evidence_verify-0.5.1-py3-none-any.whl Python 3 none any Details for agent_evidence_verify-0.5.1-py3-none-any.whl

Total release size: 1.1 MB

Release files / agent_evidence_verify-0.5.1.tar.gz

Download URL agent_evidence_verify-0.5.1.tar.gz
Size 968.8 kB
Tags Source
SHA-256 checksum
How to use checksums
c12710089e6abc559ca79b8c15e3b2982f368ec5392d66dd1dd00d835aabf3ec
BLAKE2b-256 checksum
How to use checksums
7c1b546cf0bcd2e67bf213e625bb147d580cfef296f4ba5268482c1fe666378b
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.14.7

Release files / agent_evidence_verify-0.5.1-py3-none-any.whl

Download URL agent_evidence_verify-0.5.1-py3-none-any.whl
Size 175.5 kB
Tags Python 3
SHA-256 checksum
How to use checksums
373342b9dfb7b32d1779ff2cdf4a27ade981fead9393b75a88ffbd1f729a7284
BLAKE2b-256 checksum
How to use checksums
20c4525072fbac020fa70218874ae15ae8f4dff199d1a1e08ca6db9b25603a67
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.14.7

Security for agent-evidence-verify 0.5.1

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.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A 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.

Release history Release notifications | RSS feed

This release

0.5.1 This release

2 release files No provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page