APIgentMan 🕵️‍♂️

APIgentMan is an agentic CLI tool for testing and monitoring APIs — like Postman but lighter, autonomous, and terminal-first. It can run API test collections, monitor performance, detect anomalies, and even generate new tests via an LLM.

---

✨ Features

• Run API collections (YAML format) from the CLI
• Assertions on status codes, response time, and payloads
• Performance tracking with simple anomaly detection
• LLM-powered test generation (gen-tests)
• Monitor mode for repeated runs
• Rich CLI with colored output and reports

---

📂 Project Structure

.
├── main.py              # APIgentMan CLI entrypoint
├── sample_test_app.py   # Example FastAPI app to test against
├── sample.yaml          # Sample test collection
└── README.md            # You are here

---

⚡ Setup

1. Clone & Install

git clone https://github.com/dhirajdj30/APIgentMan.git
cd apigentman
uv add fastapi groq httpx pydantic python-dotenv pyyaml rich typer uvicorn

2. (Optional) LLM Integration

Export your OpenAI key:

export OPENAI_API_KEY="sk-xxxxx"

---

▶️ Usage

1. Start the Sample API

Run the demo FastAPI server:

uvicorn sample_test_app:app --reload --port 8000

This exposes:

• GET /ping
• GET /items/{item_id}
• POST /items

2. Run Tests

Execute the sample test collection:

python main.py run sample.yaml

3. Generate Tests with LLM

python main.py gen-tests --prompt "Generate tests for a weather API with /current and /forecast endpoints"

4. Monitor Mode

python main.py monitor sample.yaml --interval 30

---

🧪 Example sample.yaml

tests:
  - name: ping test
    method: GET
    url: http://localhost:8000/ping
    expect_status: 200

  - name: get item
    method: GET
    url: http://localhost:8000/items/1
    expect_status: 200
    expect_body_contains: "item"

  - name: create item
    method: POST
    url: http://localhost:8000/items
    json: {"name": "Book"}
    expect_status: 200
    expect_body_contains: "Book"

---

🔮 Roadmap

• Export performance metrics (Prometheus/Grafana integration)
• Slack/email alerts on anomalies
• Load testing support
• Self-healing test logic
• OpenAPI → test suite auto-generation