Metadata-Version: 2.4
Name: uaal-core
Version: 0.1.0
Summary: Universal AI Authorization Layer (UAAL): policy gating, evidence logging, and Merkle-based verification for autonomous agents
Author-email: Chandan Galani <info@lolasolution.com>
License: Apache-2.0
Project-URL: Homepage, https://github.com/LOLA0786/uaal-core
Project-URL: Repository, https://github.com/LOLA0786/uaal-core
Project-URL: Issues, https://github.com/LOLA0786/uaal-core/issues
Requires-Python: >=3.9
Description-Content-Type: text/markdown
Requires-Dist: fastapi>=0.100
Requires-Dist: pydantic>=2.0

# UAAL Core

**Universal AI Authorization Layer (UAAL)**

UAAL is a lightweight control plane for autonomous AI systems.
It enforces policy decisions *before* actions execute and produces
cryptographically verifiable evidence for every decision.

## What it does

- Authorizes AI agent actions in real time
- Blocks unsafe or non-compliant actions
- Records immutable evidence per decision
- Generates daily Merkle roots for audit integrity
- Verifies tampering independently

## Install

```bash
pip install uaal-core
```

## Minimal Usage

```python
from uaal import authorize, emit_evidence

decision = authorize(
    agent="pricing-agent",
    action="update_price",
    payload={"new_price": 120}
)

emit_evidence(decision)
```

## Verify Integrity

```python
from uaal.verify import verify_day
print(verify_day())
```

## Philosophy

UAAL is not an agent framework.
It is the **seatbelt, black box, and kill switch** for AI autonomy.

## License
Apache 2.0
