From 00f3edd9cc03f598d39a875b549d851a37e4a161 Mon Sep 17 00:00:00 2001 From: Panniantong Date: Wed, 25 Feb 2026 15:11:51 +0100 Subject: [PATCH] =?UTF-8?q?refactor:=20=E5=90=88=E5=B9=B6=20SKILL.md=20?= =?UTF-8?q?=E4=B8=BA=E5=8D=95=E4=B8=80=E6=9D=A5=E6=BA=90=EF=BC=8C=E6=9B=B4?= =?UTF-8?q?=E6=96=B0=E4=B8=BA=2012=20=E4=B8=AA=E6=B8=A0=E9=81=93?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit - 删除 agent-reach/SKILL.md（重复） - 删除 agent_reach/integrations/skill/SKILL.md（重复） - 保留并更新 agent_reach/skill/SKILL.md 为唯一来源 - 新增 Instagram/LinkedIn/Boss直聘 搜索命令 - 新增渠道配置引导流程说明 - 描述从 9+ 更新为 12+ --- agent-reach/SKILL.md | 87 ------------------------- agent_reach/integrations/skill/SKILL.md | 65 ------------------ agent_reach/skill/SKILL.md | 56 ++++++++++------ 3 files changed, 37 insertions(+), 171 deletions(-) delete mode 100644 agent-reach/SKILL.md delete mode 100644 agent_reach/integrations/skill/SKILL.md diff --git a/agent-reach/SKILL.md b/agent-reach/SKILL.md deleted file mode 100644 index bb224fc..0000000 --- a/agent-reach/SKILL.md +++ /dev/null @@ -1,87 +0,0 @@ ---- -name: agent-reach -description: > - Give your AI agent eyes to see the entire internet. Read and search across - Twitter/X, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu, RSS, and any web page - — all from a single CLI. Use when: (1) reading content from URLs (tweets, Reddit posts, - articles, videos), (2) searching across platforms (web, Twitter, Reddit, GitHub, YouTube, - Bilibili, XiaoHongShu), (3) checking channel health or updating Agent Reach. - Triggers: "search Twitter/Reddit/YouTube", "read this URL", "find posts about", - "搜索", "读取", "查一下", "看看这个链接". ---- - -# Agent Reach - -Read and search the internet across 9+ platforms via unified CLI. - -## Setup - -First check if agent-reach is installed: -```bash -agent-reach doctor -``` - -If command not found, install it: -```bash -pip install https://github.com/Panniantong/agent-reach/archive/main.zip -agent-reach install --env=auto -``` - -`install` auto-detects your environment and installs all dependencies (Node.js, mcporter, bird CLI, gh CLI). Read the output and run `agent-reach doctor` to see what's active. - -For channels that need user input, ask the user. See the full setup guide: -https://raw.githubusercontent.com/Panniantong/agent-reach/main/docs/install.md - -## Commands - -### Read any URL -```bash -agent-reach read -agent-reach read --json # structured output -``` -Handles: tweets, Reddit posts, articles, YouTube (transcripts), GitHub repos, etc. - -### Search - -```bash -agent-reach search "query" # web search (Exa) -agent-reach search-twitter "query" # Twitter/X -agent-reach search-reddit "query" # Reddit (--sub ) -agent-reach search-github "query" # GitHub (--lang ) -agent-reach search-youtube "query" # YouTube -agent-reach search-bilibili "query" # Bilibili (B站) -agent-reach search-xhs "query" # XiaoHongShu (小红书) -``` - -All search commands support `-n ` for number of results. - -### Management - -```bash -agent-reach doctor # channel status overview -agent-reach watch # quick health + update check (for scheduled tasks) -agent-reach check-update # check for new versions -``` - -### Configure channels - -```bash -agent-reach configure twitter-cookies "auth_token=xxx; ct0=yyy" -agent-reach configure proxy http://user:pass@ip:port -agent-reach configure --from-browser chrome # auto-extract cookies -``` - -## Channel Status Tiers - -- **Tier 0 (zero config):** Web, YouTube, RSS, Twitter (read-only via Jina) -- **Tier 1 (free setup):** Exa web search (mcporter required) -- **Tier 2 (user config):** Twitter search (cookie), Reddit full (proxy), GitHub (token), Bilibili (proxy), XiaoHongShu (MCP) - -Run `agent-reach doctor` to see which channels are active. - -## Tips - -- Always try `agent-reach read ` first for any URL — it auto-detects the platform -- For Twitter cookies, recommend the user install [Cookie-Editor](https://chromewebstore.google.com/detail/cookie-editor/hlkenndednhfkekhgcdicdfddnkalmdm) Chrome extension -- Reddit and Bilibili block server IPs — suggest a residential proxy (~$1/month) if on a server -- If a channel breaks, run `agent-reach doctor` to diagnose diff --git a/agent_reach/integrations/skill/SKILL.md b/agent_reach/integrations/skill/SKILL.md deleted file mode 100644 index 62fa7d5..0000000 --- a/agent_reach/integrations/skill/SKILL.md +++ /dev/null @@ -1,65 +0,0 @@ ---- -name: agent-reach -description: "Give your AI Agent eyes to see the entire internet. Search + read 10+ platforms with one install." -install: pip install git+https://github.com/Panniantong/agent-reach.git ---- - -# Agent Reach — 给 Agent 装上眼睛 👁️ - -安装后你就能搜索和阅读全网内容。 - -## 安装 - -```bash -pip install git+https://github.com/Panniantong/agent-reach.git -``` - -## 装完直接能用的（零配置） - -- `agent-reach read ` — 读取任意网页、GitHub、B站、YouTube、RSS -- `agent-reach search-github "query"` — 搜索 GitHub 仓库 -- `agent-reach doctor` — 查看所有平台状态 - -## 解锁更多功能 - -运行 `agent-reach configure` 进行交互式配置，或者读取以下指南： - -| 功能 | 指南文件 | -|------|---------| -| 全网搜索 + Reddit搜索 + Twitter搜索 | `agent_reach/guides/setup-exa.md` | -| Reddit 完整阅读 | `agent_reach/guides/setup-reddit.md` | -| Twitter 高级（时间线/线程） | `agent_reach/guides/setup-twitter.md` | -| 小红书 | `agent_reach/guides/setup-xiaohongshu.md` | -| 微信公众号 | `agent_reach/guides/setup-wechat.md` | -| 视频语音转文字 | `agent_reach/guides/setup-groq.md` | - -每个指南里有详细的步骤说明，你（Agent）可以照着做，只有需要人类操作的部分（登录、复制 key）才需要问用户。 - -## MCP Server - -如果你的 Agent 平台支持 MCP： - -```bash -pip install agent-reach[mcp] -python -m agent_reach.integrations.mcp_server -``` - -提供 8 个工具：read_url, read_batch, detect_platform, search, search_reddit, search_github, search_twitter, get_status - -## Python API - -```python -from agent_reach import AgentReach -import asyncio - -eyes = AgentReach() - -# 读取 -result = asyncio.run(eyes.read("https://github.com/openai/gpt-4")) - -# 搜索 -results = asyncio.run(eyes.search("AI agent framework")) - -# 健康检查 -print(eyes.doctor_report()) -``` diff --git a/agent_reach/skill/SKILL.md b/agent_reach/skill/SKILL.md index bb224fc..8c6a455 100644 --- a/agent_reach/skill/SKILL.md +++ b/agent_reach/skill/SKILL.md @@ -2,17 +2,21 @@ name: agent-reach description: > Give your AI agent eyes to see the entire internet. Read and search across - Twitter/X, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu, RSS, and any web page - — all from a single CLI. Use when: (1) reading content from URLs (tweets, Reddit posts, - articles, videos), (2) searching across platforms (web, Twitter, Reddit, GitHub, YouTube, - Bilibili, XiaoHongShu), (3) checking channel health or updating Agent Reach. + Twitter/X, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu, Instagram, LinkedIn, + Boss直聘, RSS, and any web page — all from a single CLI. + Use when: (1) reading content from URLs (tweets, Reddit posts, articles, videos), + (2) searching across platforms (web, Twitter, Reddit, GitHub, YouTube, Bilibili, + XiaoHongShu, Instagram, LinkedIn, Boss直聘), + (3) user asks to configure/enable a platform channel, + (4) checking channel health or updating Agent Reach. Triggers: "search Twitter/Reddit/YouTube", "read this URL", "find posts about", - "搜索", "读取", "查一下", "看看这个链接". + "搜索", "读取", "查一下", "看看这个链接", + "帮我配", "帮我添加", "帮我安装". --- # Agent Reach -Read and search the internet across 9+ platforms via unified CLI. +Read and search the internet across 12+ platforms via unified CLI. ## Setup @@ -27,10 +31,7 @@ pip install https://github.com/Panniantong/agent-reach/archive/main.zip agent-reach install --env=auto ``` -`install` auto-detects your environment and installs all dependencies (Node.js, mcporter, bird CLI, gh CLI). Read the output and run `agent-reach doctor` to see what's active. - -For channels that need user input, ask the user. See the full setup guide: -https://raw.githubusercontent.com/Panniantong/agent-reach/main/docs/install.md +`install` auto-detects your environment and installs core dependencies (Node.js, mcporter, bird CLI, gh CLI, instaloader). Read the output and run `agent-reach doctor` to see what's active. ## Commands @@ -39,7 +40,7 @@ https://raw.githubusercontent.com/Panniantong/agent-reach/main/docs/install.md agent-reach read agent-reach read --json # structured output ``` -Handles: tweets, Reddit posts, articles, YouTube (transcripts), GitHub repos, etc. +Handles: tweets, Reddit posts, articles, YouTube/Bilibili (transcripts), GitHub repos, Instagram posts, LinkedIn profiles, Boss直聘 jobs, XiaoHongShu notes, RSS feeds, and any web page. ### Search @@ -51,6 +52,9 @@ agent-reach search-github "query" # GitHub (--lang ) agent-reach search-youtube "query" # YouTube agent-reach search-bilibili "query" # Bilibili (B站) agent-reach search-xhs "query" # XiaoHongShu (小红书) +agent-reach search-instagram "query" # Instagram +agent-reach search-linkedin "query" # LinkedIn +agent-reach search-bosszhipin "query" # Boss直聘 ``` All search commands support `-n ` for number of results. @@ -67,21 +71,35 @@ agent-reach check-update # check for new versions ```bash agent-reach configure twitter-cookies "auth_token=xxx; ct0=yyy" +agent-reach configure instagram-cookies "sessionid=xxx; csrftoken=yyy; ..." agent-reach configure proxy http://user:pass@ip:port -agent-reach configure --from-browser chrome # auto-extract cookies +agent-reach configure --from-browser chrome # auto-extract cookies from local browser ``` -## Channel Status Tiers +## Configuring a channel ("帮我配 XXX") -- **Tier 0 (zero config):** Web, YouTube, RSS, Twitter (read-only via Jina) -- **Tier 1 (free setup):** Exa web search (mcporter required) -- **Tier 2 (user config):** Twitter search (cookie), Reddit full (proxy), GitHub (token), Bilibili (proxy), XiaoHongShu (MCP) +All channels follow the same flow. When a user asks to configure/enable any channel: -Run `agent-reach doctor` to see which channels are active. +1. Run `agent-reach doctor` +2. Find the channel in the output — it shows the current status (✅/⚠️/⬜) and **what to do next** +3. Execute whatever you can automatically (install packages, start services, register MCP) +4. For steps that require human action (scan QR code, paste cookies, login), tell the user exactly what to do +5. After the user completes their part, run `agent-reach doctor` again to verify + +**Do NOT memorize per-channel steps.** Always rely on `doctor` output — it stays up-to-date even when backends change. + +### Common human actions + +These are things only the user can do. When `doctor` indicates one is needed, explain it clearly: + +- **Paste cookies:** User installs [Cookie-Editor](https://chromewebstore.google.com/detail/cookie-editor/hlkenndednhfkekhgcdicdfddnkalmdm) Chrome extension → goes to the website → exports Header String → sends it to you → you run `agent-reach configure -cookies "..."` +- **Scan QR code:** User opens the URL shown in `doctor` output on their phone/browser and scans with the platform's app +- **Browser login:** Some MCP services need a one-time browser login; on servers without a display, user may need VNC +- **Proxy:** Reddit/Bilibili/XiaoHongShu block server IPs — suggest a residential proxy (~$1/month) if on a server ## Tips - Always try `agent-reach read ` first for any URL — it auto-detects the platform -- For Twitter cookies, recommend the user install [Cookie-Editor](https://chromewebstore.google.com/detail/cookie-editor/hlkenndednhfkekhgcdicdfddnkalmdm) Chrome extension -- Reddit and Bilibili block server IPs — suggest a residential proxy (~$1/month) if on a server +- If a channel is ⬜ but the user hasn't asked for it, don't push — let them opt in - If a channel breaks, run `agent-reach doctor` to diagnose +- LinkedIn and Boss直聘 have Jina Reader fallback even without full setup