quark-agent
Git for AI agents. Cognitive fingerprinting CLI for Quark-AI.
pip install quark-agent
Quickstart
quark init # initialize repo
quark login # authenticate
quark snapshot --agent gpt-4o \
--prompt "You are a customer support agent" # capture a fossil
quark list # list fossils
quark diff abc12345 def67890 # compare two versions
quark restore abc12345 --to behavior_v1.md # export behavior card
How it works
Each quark snapshot runs your agent description through 7 calibrated cognitive probes and captures the behavioral fingerprint — its reasoning patterns, invariants, and stability score — as an immutable fossil.
quark snapshot --agent my_agent --prompt "…"
│
├─ 7 probes: direct · rephrased · abstract · concrete
│ adversarial · minimal · contextual
│
├─ Embedding matrix (7 × 256) → cosine similarity
│
├─ stability_score σ (0 → 1, higher = more consistent)
├─ behavioral_invariants (persistent patterns)
├─ cognitive_signature (fingerprint string)
└─ sensitive_zones (instability hotspots)
Commands
| Command | Description |
|---|---|
quark init |
Initialize .quark/config.toml in current directory |
quark login |
Authenticate (saves token to config) |
quark snapshot |
Capture a cognitive fossil |
quark list |
List fossils |
quark diff A B |
Compare two fossils — drift, stability delta, invariant changes |
quark restore ID |
Export behavior card (markdown / prompt / json) |
CI integration
# .github/workflows/quark.yml
- name: Snapshot agent
env:
QUARK_TOKEN: ${{ secrets.QUARK_TOKEN }}
QUARK_API_URL: https://quark-ai.cordee.ovh
run: |
pip install quark-agent
quark snapshot \
--agent "${{ github.repository }}" \
--prompt-file system_prompt.txt \
--tag "${{ github.ref_name }}" \
--json > fossil.json
Options
All options can be set via env vars:
| Env var | Description |
|---|---|
QUARK_TOKEN |
JWT auth token (overrides config file) |
QUARK_API_URL |
API base URL (default: https://quark-ai.cordee.ovh) |
Built on Quark-AI — infrastructure de fossiles cognitifs.
Metadata
Release files for quark-agent 0.1.0
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 | Actions |
|---|---|---|---|---|
| quark_agent-0.1.0.tar.gz | 11.7 kB | Without provenance | Details for quark_agent-0.1.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| quark_agent-0.1.0-py3-none-any.whl | Python 3 | none | any | Details for quark_agent-0.1.0-py3-none-any.whl |
Total release size: 30.4 kB
Release files / quark_agent-0.1.0.tar.gz
| Download URL | quark_agent-0.1.0.tar.gz |
|---|---|
| Size | 11.7 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
f07bfd2a147257edfb26922c373c22cd06d29f1a492861b47f3cf530ecdb3797
|
|
BLAKE2b-256 checksum How to use checksums |
e1324d61fe9ad9ae481fc38dfe09f40d8c3ba5a26d8cdd93fef27ee7ca47c5d2
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.12.3
|
Release files / quark_agent-0.1.0-py3-none-any.whl
| Download URL | quark_agent-0.1.0-py3-none-any.whl |
|---|---|
| Size | 18.7 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
aacc55481584dfa867c877a14833e9feff9335dbd52215ac202b30667d7dea4c
|
|
BLAKE2b-256 checksum How to use checksums |
53b03a3715b0fed07bf9681cd5b02f0774edbb27dd0e39da8a5feefe0c47b65e
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.12.3
|
Security for quark-agent 0.1.0
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.
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.