skills.rest
qqlcx5
qqlcx5COMMUNITY
Explore all skills by this author →
WebsiteGitHub

agent-reach

Routes internet research requests across 15 platforms using multi-backend CLI tools and APIs.

Updated Aug 15, 2026
Search Tags:#social media#github#web research#content retrieval#subtitle extraction#multi-platform search#cli routing
One-click install
npx skills add https://github.com/qqlcx5/skills-hub --skill agent-reach-qqlcx5
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-reach
Source: https://github.com/qqlcx5/skills-hub/tree/main/skills/agent-reach
Command: npx skills add https://github.com/qqlcx5/skills-hub --skill agent-reach-qqlcx5

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Gathering information from the internet requires juggling dozens of platforms—search engines, social media, developer tools, video sites—each with different APIs, authentication requirements, and CLI tools. This Skill acts as a unified router that picks the right backend for each platform and handles login-state checks, retries, and fallbacks automatically. ## Core Features & Use Cases - Multi-Platform Routing: Covers 15 platforms including Exa search, Twitter/X, Xiaohongshu, Bilibili, Reddit, LinkedIn, GitHub, YouTube, V2EX, Facebook, Instagram, RSS, and Xiaoyuzhou podcasts, with per-platform backend selection via agent-reach doctor --json. - Zero-Config Channels: Six channels work immediately—Exa web search, Jina Reader for webpages, GitHub CLI, yt-dlp for YouTube subtitles, V2EX public API, and bili-cli for Bilibili. - Authenticated Backends with Retry Chains: For login-required platforms, routes between OpenCLI (browser session reuse), platform-specific CLIs, and MCP servers, with documented fallback sequences when commands fail. - Use Case: A user asks to research how people are discussing a new AI product. The agent runs Exa for web coverage, checks Twitter and Reddit for discussion, scans Xiaohongshu and Bilibili for Chinese-language reactions, then compiles a cross-platform summary. ## Quick Start Ask the agent to research a topic across the internet, for example: "Use agent-reach to research what people are saying about the latest Claude model on Twitter, Reddit, and Xiaohongshu."

Frequently Asked Questions about agent-reach

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I search Twitter, Reddit, and Xiaohongshu from one tool?▼

Run agent-reach doctor --json first to see which backend is active for each platform, then use the corresponding command group such as twitter search, opencli reddit search, or opencli xiaohongshu search. Login-required platforms need cookies configured beforehand.

How to extract subtitles from YouTube and Bilibili videos?▼

For YouTube, use yt-dlp with --write-sub --skip-download to fetch subtitles without downloading video. For Bilibili, do not use yt-dlp since it is blocked; use opencli bilibili subtitle BVxxx or bili-cli for metadata instead.

Does agent-reach work without configuring any API keys?▼

Yes, six channels work with zero configuration: Exa web search, Jina Reader for webpages, GitHub CLI, yt-dlp for YouTube, the V2EX public API, and bili-cli for Bilibili. Other platforms require cookies or login state.

Why does Twitter search fail and how do I fix it?▼

Twitter frequently changes its GraphQL endpoints, causing intermittent 404 errors on search. Retry once, upgrade twitter-cli with pipx, then fall back to opencli twitter search using your browser session, or use stable commands like twitter feed instead.

Can agent-reach post, comment, or like content on social platforms?▼

No, the skill is read-only by design and explicitly excludes write operations like posting, commenting, or liking. It only retrieves content from the internet and does not handle report writing or data analysis either.

Related Skills

Explore More Education & Research Skills→
xianmingyaoxianmingyao
1

agent-reach

Automate installation and configuration of upstream tools across 13+ platforms.

Community
Advanced
gooseworks-aigooseworks-ai
1.1k

twitter-scraper

Retrieve and deduplicate Twitter/X posts matching a query with date filters.

Official
Intermediate
RonanCodesRonanCodes

transcribe

Transcribe audio or video into timestamped text using Whisper backends.

Community
Advanced
opensquillaopensquilla
6.8k

deep-research

Conducts multi-round research with evidence tracking and citation-tagged report synthesis.

Official
Advanced