Receipt-native AI safety toolkit
Project description
Assay
Receipt-native AI safety toolkit. Build, sign, and verify Proof Packs that prove what your AI system actually did.
Install
pip install assay-ai
60-Second Demo
assay demo-pack
Creates synthetic receipts, builds signed Proof Packs, and verifies them. No API key, no configuration required.
Core Flow
# 1. Scan your project for uninstrumented LLM calls
assay scan .
# 2. Instrument your code (one line)
# from assay.integrations.openai import patch; patch()
# 3. Run your code through Assay
assay run -c receipt_completeness -c guardian_enforcement -- python my_agent.py
# 4. Verify the output
assay verify-pack ./proof_pack_*/
# 5. Gate in CI
assay verify-pack ./proof_pack_*/ --require-claim-pass
What Assay Does
Assay is a verification layer for AI systems. It does two jobs:
- Courthouse job: prove structural integrity of evidence (signatures, hashes, required files)
- Laboratory job: test behavioral claims against that evidence (RunCards)
These are deliberately orthogonal: integrity can pass while claims fail. That's the honesty property.
Commands
| Command | Purpose |
|---|---|
assay demo-pack |
Generate demo packs (no config needed) |
assay scan |
Find uninstrumented LLM call sites |
assay run |
Wrap command, collect receipts, build signed pack |
assay verify-pack |
Verify a Proof Pack (integrity + claims) |
assay lock write |
Freeze verification contract to lockfile |
assay lock check |
Validate lockfile against current card definitions |
assay doctor |
Preflight check: is Assay ready here? |
Exit Codes
- 0 -- integrity PASS and claims PASS
- 1 -- integrity PASS, claims FAIL
- 2 -- integrity FAIL or lock mismatch
Documentation
Related Repos
| Repo | Purpose |
|---|---|
| assay | Core CLI, SDK, conformance corpus (this repo) |
| assay-verify-action | GitHub Action for CI verification |
| assay-ledger | Public transparency ledger |
License
Apache-2.0
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file assay_ai-1.1.0.tar.gz.
File metadata
- Download URL: assay_ai-1.1.0.tar.gz
- Upload date:
- Size: 92.8 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.11.14
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
48af3f7316614524415c6f8788d177062302f693f233fd51b9deccebef52a858
|
|
| MD5 |
5ef3c06eb520e2fa518fcc2d32d030b0
|
|
| BLAKE2b-256 |
13cf6a06f06f5842725ea9c033beab1e854fe748d8af3d3580184e0e64740960
|
File details
Details for the file assay_ai-1.1.0-py3-none-any.whl.
File metadata
- Download URL: assay_ai-1.1.0-py3-none-any.whl
- Upload date:
- Size: 107.0 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.11.14
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
251a475ff573115786897ff72e936dc381c4c56162a7ab0ec0ee411b8c70db82
|
|
| MD5 |
dd65dacd319776877c015ba7a94c7cb1
|
|
| BLAKE2b-256 |
899f0ec3dd0ed91bc37b0876d34d4fe7d4630e2fb486210919aa3bf35a59c2c4
|