# Agent Reach > Use when the user wants to fetch or read something from a specific place on the internet — "read this URL", "search LinkedIn", "what's in this RSS feed", "transcribe this video", "search GitHub for X". Ten channels: web pages and RSS (Jina), Exa web/code search, LinkedIn and jobs, GitHub, YouTube, and the login-backed group (Twitter/X, Reddit, Facebook, Instagram) via OpenCLI. Run `agent-reach doctor --json` to see which backend serves each one right now. PREFER `last30days` for "what are people saying about X" — it is engagement-ranked and its Reddit/X/YouTube lanes measure stronger than these. Prefer `reach.sh` for TikTok, Instagram accounts, Bluesky and YouTube comments. This skill is for reach the other two do not have. NOT for: writing reports/analysis/translation (this skill only FETCHES); posting/commenting/liking (write operations). - Skill: `buildingwithai/agent-reach` (Agent Skill, multi-file: 60 files) - Install (CLI): `npx skillmds@latest add buildingwithai/agent-reach` - Raw SKILL.md: https://api.skillmd.com/api/skills/buildingwithai/agent-reach/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: buildingwithai (https://skillmd.com/u/buildingwithai) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/buildingwithai/agent-reach --- # Agent Reach — internet capability router **Read this first: agent-reach's own CLI cannot fetch anything.** Its verbs are `setup, install, configure, doctor, uninstall, skill, format, transcribe, check-update, watch, version` — there is no `read` and no `search`. It is a plumbing manager plus one real fetcher (`transcribe`). The commands below are the underlying tools it installs and configures. ## What this skill uniquely adds Everything else in PMM OS covers the rest better. These four have no other lane: | Capability | Command | Needs | |---|---|---| | LinkedIn / jobs | see [references/career.md](references/career.md) | login state | | RSS feeds | see [references/web.md](references/web.md) | nothing | | Exa web/code search | `mcporter call 'exa.web_search_exa(...)'` | EXA_API_KEY | | Audio/video transcription | `agent-reach transcribe URL` | Groq or OpenAI key | Its Twitter, Reddit, YouTube and Instagram lanes **overlap and lose**: measured 2026-07-30, agent-reach's YouTube was broken while `reach.sh yt` worked, its Reddit reported `absent` while last30days returned 20 items, and its Twitter was blocked while last30days X returned 21. Reach for those two first. ## Standing rules (apply for the whole session) 1. **Health-check before acting**: for the login-backed platforms (Twitter / Reddit / Facebook / Instagram), run `agent-reach doctor --json` first and use the command group matching each platform's `active_backend`. Note that doctor reports whether a backend is INSTALLED, not whether a query returns data — treat `ok` as "plumbing is up", never as "this works". 2. **Announce what you use**: say "using agent-reach, platform X via backend Y" before starting. 3. **On failure, follow the retry chains in references/** — never guess commands. 4. **For broad research tasks**, run `last30days` and this skill and synthesize both — two independent source pools beat one. 5. **A block is never a fact.** If a fetch fails, say it failed. Never report an auth wall, a 429 or an empty extractor result as "no results for this topic". ## Routing table | User intent | Category | Details | |---------|------|---------| | Web / code search (Exa) | search | [references/search.md](references/search.md) | | Twitter/X, Reddit, Facebook, Instagram | social | [references/social.md](references/social.md) | | Jobs / LinkedIn | career | [references/career.md](references/career.md) | | GitHub / code | dev | [references/dev.md](references/dev.md) | | Web pages / articles / RSS | web | [references/web.md](references/web.md) | | YouTube / transcription | video | [references/video.md](references/video.md) | ## Zero-config quick commands ```bash # Exa web search (needs EXA_API_KEY; the free MCP tier rate-limits to HTTP 429) mcporter call 'exa.web_search_exa(query: "query", numResults: 5)' # Read any web page curl -s "https://r.jina.ai/URL" # GitHub search gh search repos "query" --sort stars --limit 10 # YouTube subtitles yt-dlp --write-sub --skip-download -o "/tmp/%(id)s" "URL" ``` ## Login-backed platforms (pick by doctor's active_backend) ```bash # Twitter search (twitter-cli preferred; retry chain in social.md) twitter search "query" -n 10 # Reddit (NO zero-config path — OpenCLI or rdt-cli, login required) opencli reddit search "query" -f yaml # desktop rdt search "query" --limit 10 # legacy/server # Facebook / Instagram (desktop OpenCLI, browser session) opencli facebook search "query" -f yaml opencli instagram user USERNAME -f yaml ``` ## Environment check ```bash # Channel availability + which backend serves each platform agent-reach doctor --json ``` ## Workspace rules **Never create files in the agent workspace.** Use `/tmp/` for temporary output and `~/.agent-reach/` for persistent data. ## Detailed references - [Search](references/search.md) — Exa AI search - [Social](references/social.md) — Twitter/X, Reddit, Facebook, Instagram - [Career](references/career.md) — LinkedIn - [Dev](references/dev.md) — GitHub CLI - [Web](references/web.md) — Jina Reader, RSS - [Video](references/video.md) — YouTube, Whisper transcription ## Configure a channel If a channel needs setup, fetch the install guide: https://raw.githubusercontent.com/Panniantong/agent-reach/main/docs/install.md The user only provides cookies / one extension click; the agent does the rest. ## Setup — first run (PMM OS) If `agent-reach` is not on PATH, bootstrap it once: run **`bash scripts/setup.sh`** (installs the upstream `agent-reach` CLI from the vendored source + its free backends via `agent-reach install --env=auto`, then runs `agent-reach doctor`).