🔍 ProbeAI — Autonomous API Reliability & Postman Suite Generator

An autonomous developer CLI agent that inspects OpenAPI specs, synthesizes multi-scenario test suites, executes real-time HTTP calls, and auto-exports ready-to-import Postman collections.

⚡ Quickstart

Option 1: Run Instantly via npx (Zero Installation Required)

npx @kamrankhan027/probe-ai@latest

Option 2: Install Globally

Using npm:

npm install -g @kamrankhan027/probe-ai

Using pnpm / yarn:

pnpm add -g @kamrankhan027/probe-ai
# or
yarn global add @kamrankhan027/probe-ai

Then run ProbeAI from any directory on your machine:

probe-ai

🧠 Supported AI Models & Providers

ProbeAI is universal and provider-agnostic. On startup, select your preferred engine:

| Provider | Supported Models | Environment Variable | |---|---|---| | 🔷 Google Gemini | gemini-3.6-flash, gemini-3.5-flash, gemini-2.5-flash, gemini-2.5-pro | GEMINI_API_KEY | | 🟢 OpenAI | gpt-5.6 (Sol/Terra/Luna), gpt-4o, gpt-4o-mini, o3-mini, gpt-oss | OPENAI_API_KEY | | 🟣 Anthropic Claude | claude-sonnet-5, claude-opus-5, claude-3-5-sonnet-latest, claude-3-5-haiku | ANTHROPIC_API_KEY | | ⚡ Groq (LPU Engine) | llama-3.3-70b-versatile, gpt-oss-120b, qwen-3.6 (Free & Ultra-Fast) | GROQ_API_KEY | | 🌐 OpenRouter | deepseek/deepseek-r1:free, qwen/qwen-2.5-coder-32b-instruct:free, mistral-large-3 | OPENROUTER_API_KEY | | 🦙 Ollama (Local) | qwen2.5-coder:7b, deepseek-r1, llama3.3, gemma-4 (100% Offline & Free) | None needed! |

💡 Custom Model Overrides: You can pass any custom model string directly via environment variables (e.g. OPENAI_MODEL=gpt-5.6 or ANTHROPIC_MODEL=claude-sonnet-5) or during the interactive startup prompt.

🌟 Key Superpowers

• 🔍 Auto-Discovery via OpenAPI: Probes any API base URL (e.g. http://localhost:8000) and parses /openapi.json to map all routes, parameters, and security schemas.
• 📦 1-Click Postman Collection Export: Eliminates 30+ minutes of manual route recreation by generating ready-to-import Postman Collection v2.1 files (reports/probe_collection.json).
• 🔐 Smart Auth Detective: Supports direct token injection (Bearer ... or X-API-Key) AND autonomous login flows (executing POST /auth/login, extracting the JWT token, and binding it to subsequent requests).
• ⚡ Real-Time HTTP Latency Profiling: Fires live requests, catches 422/404/401 edge cases, and logs latency in milliseconds.
• 📊 Standalone HTML & Terminal Reports: Generates interactive single-page HTML audit dashboards with health scores and payload replays.

💡 Example Prompts to Try

❯ inspect https://bookmark-agent-backend.onrender.com and export a Postman collection

Auto-maps all endpoints and creates reports/postman_collection.json.

❯ test all endpoints on https://bookmark-agent-backend.onrender.com and save an HTML report

Executes Happy Path, Query Filter, and Error test cases and saves reports/audit_report.html.

❯ set auth Bearer my_jwt_token_123 and test protected endpoints

Injects your authorization header into session memory for all subsequent calls.

🏗️ Repository Architecture

probe-ai/
├── cli/                            # 🚀 Native TypeScript NPM Package (@kamrankhan027/probe-ai)
│   ├── src/
│   │   ├── index.ts                # Interactive CLI loop & Modern UI
│   │   ├── wizard.ts               # Multi-Provider Startup Wizard
│   │   ├── model-factory.ts        # Universal LangChain chat model factory
│   │   ├── agent.ts                # @langchain/langgraph ReAct workflow
│   │   └── tools/                  # OpenAPI, HTTP, Auth, Postman & Report tools
│   ├── dist/
│   │   └── index.js                # ⚡ Single zero-dependency bundled executable (tsup)
│   └── package.json
├── python/                         # Python implementation (LangGraph + Rich)
│   ├── agent.py
│   └── requirements.txt
└── reports/                        # Shared output directory