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.
Thành tựu nổi bật
Những danh hiệu mới nhất của dự án
#1 Repository Của Ngày
#1 Repository Của Ngày
#1 Repository Của Ngày
👁️ Agent Reach: Give Your AI Agent Eyes on the Internet
One CLI, zero API fees — install once and your agent can read Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu, and more.
Agent Reach is a capability layer that gives AI agents the ability to see and interact with the internet. It handles tool selection, installation, health checks, and routing so you don't have to. Compatible with Claude Code, OpenClaw, Cursor, Windsurf, and any agent that can run shell commands. Install it once with a single command, and your agent can research the web, read social media, extract video subtitles, and subscribe to RSS feeds — all without paying for APIs.
✨ Highlights
- Zero-config web reading: Read any webpage via Jina Reader — no setup needed
- YouTube subtitle extraction: Extract subtitles and search videos with yt-dlp
- Full-net semantic search: Exa-powered semantic search across the web, free with no API key
- Multi-platform social access: Twitter/X, Reddit, Bilibili, XiaoHongShu, LinkedIn, V2EX, Xueqiu — some zero-config, some with simple login setup
- Auto-failover routing: Multiple backends per platform; if one access method breaks, switch to another seamlessly
- Self-diagnostic: Run
agent-reach doctorto check which channels work and which need attention - Cookie-based auth: Use Chrome Cookie-Editor extension to export cookies for platforms that need login — cookies stay local, never uploaded
- RSS support: Read any RSS/Atom feed with feedparser
💡 Why it matters
Agent Reach solves a real pain point: AI agents can write code and manage projects, but ask them to look something up online and they're blind. Every platform has its own hurdles — paid APIs, login walls, anti-bot blocks, data cleaning. Agent Reach turns this into a one-line install that your agent handles itself.
What sets it apart is that it's a capability layer, not just another tool. It doesn't wrap individual readers — it handles selection, installation, diagnostics, and routing. When a platform blocks one access method (like yt-dlp being blocked on Bilibili), Agent Reach switches to another backend automatically, and the user never notices. It's the difference between handing your agent a pile of broken tools and handing it a fully stocked, self-healing toolkit.
#AI #AgentInfrastructure #OpenSource #CLI #TiniX
Dự án tương tự
Khám phá các dự án và mô hình AI khác có sự tương đồng.
Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.
Production-grade engineering skills for AI coding agents.
A curated list of awesome Claude Skills, resources, and tools for customizing Claude AI workflows
Google Workspace CLI — one command-line tool for Drive, Gmail, Calendar, Sheets, Docs, Chat, Admin, and more. Dynamically built from Google Discovery Service. Includes AI agent skills.
Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or self-hosted, so teams move from prototype to production without rebuilding the stack.
Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store.
Agent skills that give coding agents taste: art direction, reference images and a browser check instead of generic AI pages
Reverse engineer anything with agents, from app behavior down to native binaries.