Search + Read across 16+ platforms including Twitter/X, Reddit, YouTube Transcripts, GitHub, Bilibili, and RSS feeds. Built-in MCP server for Claude Code, Cursor, and custom agent workflows.
npx skills add https://github.com/panniantong/agent-reach --skill agent-reach
Select a channel below, enter a URL or search term, and test real-time data retrieval.
Click Run Reach above or choose a sample to inspect real-time content extraction.
Automated fallback architecture ensures your agents maintain persistent visibility.
Extract clean video metadata, auto-generated subtitles, chapters, and full multi-language transcripts without video downloading.
Read full post threads, nested comment discussions, author metadata, and search subreddit posts without OAuth limits.
Deep-read repository READMEs, code file trees, commit history, pull requests, and public issues via clean structured JSON.
Transform raw HTML into distilled, clean Markdown text. Strips ads, navigation clutter, cookie banners, and scripts.
Track blog updates, newsletter dispatches, and news syndication feeds with instant XML/JSON normalization.
Read public tweets, thread progressions, quote posts, and user profiles with automated multi-backend routing.
Standardized protocol integration for Claude Code, Cursor, OpenClaw, and Windsurf.
# Install Agent Reach Skill in Claude Code:
npx skills add https://github.com/panniantong/agent-reach --skill agent-reach
# Or start native MCP server:
agent-reach mcpLive health verification across core scraping and extraction routes.