Summary

ZAPI by Adopt AI is an open-source Python library designed for automated network traffic and API call capture from web applications. It targets developers and researchers building AI agents or LLM training datasets, offering a streamlined approach to API discovery and understanding web application backends. The primary benefit is simplifying the process of generating structured API data for AI and development workflows.

How It Works

The library utilizes Playwright to launch browsers, navigate web applications, and capture all network traffic, saving it in HAR (HTTP Archive) format. It then employs sophisticated HAR analysis to filter out non-API relevant requests (like static assets) and provides cost and time estimations for processing. Crucially, ZAPI supports "Bring Your Own Key" (BYOK) for various LLM providers (OpenAI, Anthropic, Google, Groq) to enable enhanced, zero-shot API discovery and documentation generation.

Quick Start & Requirements

Installation involves pip install -r requirements.txt and playwright install. Prerequisites include Python 3.9+, Playwright 1.40.0+, and cryptography 41.0.0+. Environment setup requires a .env file with LLM provider details, API keys, and credentials (ADOPT_CLIENT_ID, ADOPT_SECRET_KEY) obtained from app.adopt.ai. A demo.py script is available for testing installation.

Highlighted Details

• Zero-Shot API Discovery: Automatically identifies and documents APIs by analyzing captured network traffic.
• HAR Analysis & Cost Estimation: Provides detailed statistics on captured traffic, including API relevance, unique domains, estimated processing cost, and time.
• LangChain Integration: Seamlessly converts discovered APIs into LangChain tools, facilitating agent development.
• BYOK LLM Support: Integrates with OpenAI, Anthropic, Google, and Groq, securing provided API keys.

Maintenance & Community

No specific details regarding maintainers, community channels (like Discord/Slack), or project roadmap were found in the provided README content.

Licensing & Compatibility

The provided README content does not specify a software license. This absence makes it difficult to assess compatibility for commercial use or closed-source linking without further clarification.

Limitations & Caveats

Full functionality, particularly enhanced API discovery, requires obtaining and configuring API keys for supported LLM providers and specific credentials from app.adopt.ai. The project relies on external services for authentication and discovery.