## https://sploitus.com/exploit?id=97F1E034-3D73-5D5B-86DA-90831A594737
# Pentest Agent
AI-powered penetration testing agent using Claude as the reasoning engine.
> **IMPORTANT**: Only use this against applications you own or have explicit
> authorization to test. This is for educational and authorized security
> testing purposes only.
## Quick Start
### 1. Install dependencies
```bash
cd pentest-agent
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
```
### 2. Set your API key
```bash
cp .env.example .env
# Edit .env and add your Anthropic API key
```
### 3. Start a vulnerable target (DVWA)
```bash
docker compose up -d
# Wait ~30 seconds, then visit http://localhost:8080
# Login: admin / password
# Go to http://localhost:8080/setup.php and click "Create / Reset Database"
```
### 4. Run the agent
```bash
# Full scan
python main.py --target http://localhost:8080
# Specific task
python main.py --target http://localhost:8080 --task "Find SQL injection in the login form"
# Fewer iterations (faster, less thorough)
python main.py --target http://localhost:8080 --max-iterations 10
```
## Architecture
```
main.py โ CLI entry point
agent.py โ ReAct loop (Reason โ Act โ Observe)
tools.py โ HTTP tools the agent can call
validator.py โ Deterministic validation (zero false positives)
```
## Next Steps
- Add more tools (Playwright browser, sqlmap integration)
- Build specialized agents (SQLi agent, XSS agent, etc.)
- Add a coordinator for multi-agent orchestration
- Integrate the validator into the agent loop
- Add reporting (HTML/JSON output)