Tools
Breach Intel Packages
Distribution packages for Breach Intel — install scripts, SDK wheels, OpenClaw plugin, and Docker files
Install
pip install breach-intel-client
README
# Breach Intel — Distribution Packages
Pre-built packages and installers for **Breach Intel**, the audit layer for AI agents.
> Policy agent infrastructure that attaches to any AI agent and produces tamper-evident breach audit logs — automatically. Zero code changes.
## Quick Install
```bash
# One-line install (recommended)
./install.sh
# SDK only (no Docker)
./install.sh --sdk-only
# Or install directly from PyPI
pip install breach-intel-client
```
## What's Included
### `install.sh`
Automated installer that handles everything:
- Detects OS (macOS / Linux / Windows)
- Checks prerequisites (Python 3.9+, Docker)
- Installs the Python SDK (`breach-intel-client`)
- Sets up the persistent auto-instrumentation hook
- Starts the Docker stack (policy agent + PostgreSQL)
- Registers your agent and prints the dashboard URL
### `sdk/`
Pre-built Python SDK packages (also available on [PyPI](https://pypi.org/project/breach-intel-client/)):
| File | Version | Description |
|------|---------|-------------|
| `breach_intel_client-0.3.1-py3-none-any.whl` | 0.3.1 | Latest wheel (recommended) |
| `breach_intel_client-0.3.1.tar.gz` | 0.3.1 | Latest source distribution |
| `breach_intel_client-0.1.0-py3-none-any.whl` | 0.1.0 | Legacy wheel |
| `breach_intel_client-0.1.0.tar.gz` | 0.1.0 | Legacy source distribution |
Install from local file:
```bash
pip install sdk/breach_intel_client-0.3.1-py3-none-any.whl
```
### `openclaw-plugin/`
OpenClaw integration plugin for real-time breach detection within the OpenClaw AI agent framework:
| File | Description |
|------|-------------|
| `index.ts` | Plugin entry point |
| `handler.ts` | Event handler (intercepts LLM messages) |
| `openclaw.plugin.json` | Plugin manifest |
| `install.sh` | Plugin-specific installer |
| `HOOK.md` | Integration documentation |
### `docker-compose.yml` + `Dockerfile`
Docker stack for running the policy agent server:
```bash
docker compose up --build
```
Starts:
- **Policy Agent** (FastAPI) on port 8080
- **PostgreSQL 16** for breach record storage
- **Dashboard** at `http://localhost:8080/dashboard/`
## Supported Platforms
| Platform | install.sh | SDK | OpenClaw Plugin | Docker |
|----------|-----------|-----|-----------------|--------|
| macOS | ✓ | ✓ | ✓ | ✓ |
| Linux | ✓ | ✓ | ✓ | ✓ |
| Windows | ✓ | ✓ | ✓ | ✓ |
## Requirements
- Python 3.9+
- Docker (for full install, not needed for SDK-only)
- pip
## Verify Installation
```bash
breach-intel doctor
```
## Links
- **Website**: [parthamehtaorg.github.io/breach-intel-site](https://parthamehtaorg.github.io/breach-intel-site/)
- **PyPI**: [pypi.org/project/breach-intel-client](https://pypi.org/project/breach-intel-client/)
- **Documentation**: [Docs](https://parthamehtaorg.github.io/breach-intel-site/docs.html)
## Version
Current: **v0.3.1**
## Changelog
### v0.3.1
- Python 3.9 compatibility fix (`Optional[str]` instead of `str | None`)
- Auto-register agents with policy agent on attach
- Dashboard improvements
### v0.3.0
- Healthcare vertical (14 breach types, HIPAA taxonomy)
- Pharma vertical (14 breach types, FDA 21-CFR-11)
- Agentic deep analysis (LangGraph + Claude)
### v0.2.0
- OpenClaw plugin integration
- Webhook alerts (Slack, Discord, generic)
- Auto-scaling support
### v0.1.0
- Initial release
- Fintech vertical (12 breach types)
- SHA-256 hash chain
- Live dashboard
- Auto-instrumentation (OpenAI, Anthropic, LangChain)
## License
MIT
tools
Comments
Sign in to leave a comment