Skip to main content

Proofbeak

Check whether a service meets your requirements. Decide with evidence.

PyPI · Documentation

Proofbeak is a self-hosted acceptance-testing and evidence service for developers and AI agents. Define the behavior your application needs, run the same test against configured services, and inspect independently observed results before choosing an integration or release.

For example: write data through an API, read it independently, and check that the returned fields match your requirements. Repeat with different data or candidate services; compare the result with a previous release. Proofbeak records what happened and whether cleanup succeeded.

You define scenarios through the visual editor or JSON using connector operations, test data and acceptance conditions. New cases and data combinations do not require changes to this package.

Start in the browser

Use Python 3.11+ on Linux or macOS, in a virtual environment:

python3 -m venv .venv
. .venv/bin/activate
python -m pip install 'proofbeak[agents,connectors]==0.8.0'
proofbeak studio

Open the displayed local address and sign in with the access key printed by the launcher. Choose Try a local example, edit its values, then Validate & preview and Run test package. The example uses actual disposable SQLite data. No cloud account, model or external database is needed. Configuration, results and signing identity persist in your private local workspace.

The workspace guide covers your own service profiles, visual/JSON authoring, safe retries, history, comparisons and exports.

What you can do

Capability Use
Guided authoring Choose configured operations and independent observers; specify your expected results.
Suites and parameter matrices Run multiple scenarios and bounded data combinations against compatible services.
Release comparisons Identify per-condition regressions, improvements and changed observations in compatible historical runs.
CI gates and reports Export JSON, HTML, JUnit and Markdown; fail CI on rejected or incomplete evidence.
Web, CLI, REST and MCP Use the same execution and evidence service from a browser, automation or an external agent.
Connector SDK Install trusted independent Python connector packages for additional protocols.
Signed evidence Export Ed25519-signed records and verify content offline against an independently trusted public key.

Compose tests over configured REST APIs, MCP tools, SQLite/PostgreSQL transactions, Redis Streams and HTTP search/model endpoints. Existing storage and calendar adapters remain available. Provider credentials and executable bindings belong to the operator. Additional protocols require a connector; business conditions remain yours.

Use the CLI

The core CLI has no third-party Python dependencies. This suite tests three values against actual disposable local files:

python -m pip install proofbeak==0.8.0
proofbeak suite example --output acceptance.json
proofbeak suite validate acceptance.json
proofbeak suite run acceptance.json --key first-suite \
  --output reports/suite.json --html reports/suite.html \
  --junit reports/junit.xml --summary reports/summary.md
proofbeak gate reports/suite.json

Edit the definition for your requirement and use a new key for an intentional new run. Repeating an unchanged request and key retrieves the original execution without replaying writes.

Read the result

Result Meaning
pass Recorded observations satisfy your acceptance conditions.
fail An observed value violates a condition.
inconclusive Missing observations or an operational issue prevent a conclusion.

Cleanup is reported separately. The CI gate requires actual observations, passing acceptance and clean resources. A valid signature establishes record integrity and signer identity for the chosen key; it does not certify a provider or prove it still behaves the same way.

Deploy and extend

Run directly in Python, in a container or on standard Kubernetes/Helm. The supported deployment uses one coordinator with SQLite state, tenant-bound access, bounded workers and explicit recovery. OpenShift is optional. Remote SSO, multi-DC operation and a shared public evidence catalog are outside this release. Technical qualification does not establish human adoption or an improvement in external-agent decision quality.

Contributing · Architecture · Apache-2.0

Metadata

Release files for proofbeak 0.8.0

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

Source distribution (sdist)

Source distribution for proofbeak 0.8.0
File Size Uploaded Provenance Actions
proofbeak-0.8.0.tar.gz 339.7 kB With provenance Details for proofbeak-0.8.0.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for proofbeak 0.8.0
File Interpreter ABI Platform Actions
proofbeak-0.8.0-py3-none-any.whl Python 3 none any Details for proofbeak-0.8.0-py3-none-any.whl

Total release size: 503.6 kB

Release files / proofbeak-0.8.0.tar.gz

Download URL proofbeak-0.8.0.tar.gz
Size 339.7 kB
Tags Source
SHA-256 checksum
How to use checksums
01ab39ec645257fa39d3af5d50675d377c0872488df12bc0a211ad90dfe4345d
BLAKE2b-256 checksum
How to use checksums
e20846e97530f749be1110d91c8e9167737c86aee6cde8d20bdb8ca7b9886bb3
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

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 Nov 13, 2026.

Transparency log

Release files / proofbeak-0.8.0-py3-none-any.whl

Download URL proofbeak-0.8.0-py3-none-any.whl
Size 163.9 kB
Tags Python 3
SHA-256 checksum
How to use checksums
9e20e17aa4951c011e7e737e034fe544f71028c91262794916aea127570407dc
BLAKE2b-256 checksum
How to use checksums
89043632520245dd9602cee11e52735127dbbc38f81bd37b45a0e5a3d1fcbac5
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

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 Nov 13, 2026.

Transparency log

Security for proofbeak 0.8.0

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

Both files in this release have been uploaded with provenance metadata.

Select from 2 files

Attestation for proofbeak-0.8.0.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 Nov 13, 2026.

Transparency log

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.

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