ToolNew

Agent Reach

Tool : A precise action the AI can call · requires any-agent

Gives an agent read and search access to the web, YouTube, Reddit, X, GitHub, LinkedIn, and a dozen other platforms, without a paid API key.

Very practical for agent-based monitoring, but relies on your session cookies and scraping: risk of account blocking, and platform terms of service should be reviewed before professional use.

Official site GitHub · 85,081 ★ Docs

Overall score

7.6 /10

Security: one caveat, see the card

Utility

7.0

Reliability

10.0

Security

One caveat

Ease

5 min · Comfortable

Security reflects where data goes, requested permissions and incidents on record at the verification date. It is not a guarantee of zero risk: check the deployment mode and the connections you enable.

What it does, concretely

Agent Reach solves a concrete problem: AI agents can write code and manage files, but they are blocked as soon as they need to read a web page, search Reddit discussions, or extract subtitles from a YouTube video. Each platform imposes its own barriers: paid APIs, IP blocks, mandatory login, and raw data that is hard to process.

The tool works as a capability layer installed with a single instruction given to your agent. The agent automatically downloads and configures a set of open-source tools (yt-dlp for YouTube, gh CLI for GitHub, feedparser for RSS feeds, Jina Reader for web pages, etc.) and registers a SKILL.md file that tells the agent what to call depending on the target platform. For platforms requiring authentication (X, Reddit, LinkedIn, Instagram, Facebook, Xiaohongshu), additional agent-guided configuration unlocks access via session cookies or browser automation.

Agent Reach is aimed at developers and advanced users who run AI agents from the command line and want to give them structured web access without assembling each component themselves. The agent-reach doctor command lets you check at any time which backend is active for each platform and get repair instructions.

The tool is not suited for intensive professional use on main accounts: relying on session cookies and scraping creates account-blocking risks and may conflict with the terms of service of the platforms involved. It is also not intended for users who are not comfortable with a terminal.

What it offers

·Web page reading via Jina Reader, without an API key, for any public content.
·YouTube subtitle extraction and video search via yt-dlp, without authentication.
·Read and search access to GitHub (public repositories without configuration, private repositories and actions after login via gh CLI).
·Semantic web search via Exa, integrated through MCP without a required API key.
·RSS and Atom feed reading via feedparser, without configuration.
·Access to X, Reddit, LinkedIn, Instagram, Facebook, and Xiaohongshu after configuring session cookies or browser automation (OpenCLI).
·Bilibili search and reading via bili-cli without login, with subtitles available after configuration.
·agent-reach doctor command to diagnose the status of each channel and get repair instructions.

Our tips

Run agent-reach doctor after installation to immediately check which channels are working and which ones need additional configuration.
For platforms requiring login, use a dedicated secondary account rather than your main account to limit the consequences of a potential block.
Use agent-reach install --dry-run to preview all operations before they are executed on your machine.
Only use --system when you explicitly accept that the tool will modify your system environment; without this flag, installation is limited to an environment check.
To update channels without reinstalling, give your agent the update instruction pointing to the update.md file in the repository.

Watch out

Using session cookies for X, Reddit, LinkedIn, Instagram, Facebook, and Xiaohongshu exposes accounts to a risk of blocking or suspension by the platforms, which may detect calls not originating from a normal browser.
Reddit offers no unauthenticated path: the anonymous interface is blocked and the official API requires approval, making this channel unusable without prior configuration.

In practice

Price: Free
Commercial use: Commercial use: yes
Available on: cli, skill
Host: any-agent
Licence: MIT
Other hosts: claude-code, cursor, codex, any-agent
FR interface: No

Security and data

Rating: Security: one caveat, see the card
Your data: Data stays local
Maintenance: Active
Last release: 11 June 2026

Links and repository checked on 23 September 2026 · added on 23 September 2026

How to install Agent Reach

1. Copy the following instruction and give it to your AI agent (Claude Code, Cursor, Windsurf, etc.): "Help me install Agent Reach: https://raw.githubusercontent.com/Panniantong/agent-reach/main/docs/install.md" 2. The agent downloads and runs the installation process; without the --system flag, it only checks the environment without modifying the system. 3. To allow full installation of system dependencies, explicitly add --system to the install command. 4. Once installation is complete, run agent-reach doctor to check the status of each channel. 5. For platforms requiring login (X, Reddit, etc.), tell your agent "Help me configure [platform name]" and follow the guided steps; use a dedicated secondary account.

https://raw.githubusercontent.com/Panniantong/agent-reach/main/docs/install.md

Official documentation ↗

In the same category

Whole category Research & monitoring →