Panniantong/Agent-Reach
Give your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
Provides browser automation and web scraping tools via a unified CLI that abstracts platform-specific challenges—integrating yt-dlp for video transcription, bird for Twitter access, Jina Reader for web parsing, and MCP servers for social platforms—while storing credentials locally and supporting diagnostic checks via `agent-reach doctor`. Designed as a pluggable scaffolding where each platform backend is independently replaceable, enabling AI agents running on any framework (Claude Code, Cursor, OpenClaw) to read restricted content without API fees or custom configuration per tool.
8,832 stars. Actively maintained with 96 commits in the last 30 days.
Stars
8,832
Forks
636
Language
Python
License
MIT
Category
Last pushed
Mar 13, 2026
Commits (30d)
96
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/perception/Panniantong/Agent-Reach"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.