mcpField
A
agent-reach

agent-reach

github.com/ootto-aiProductivity32968 installsUpdated 2w ago
Repository

About agent-reach

agent-reach is an internet-capability routing skill for agents, directing research across search, social, web, career, development, and video sources. It selects platform-specific backends through reference guides and includes environment checks and retry procedures.

Key features

  1. 01Routes across 13 platforms and multiple backends
  2. 02Supports search, web reading, and code search
  3. 03Covers social media and career sources
  4. 04Supports video and podcast transcript research
  5. 05Checks available backends with doctor

Use cases

  1. 01Research a topic across platforms
  2. 02Find social discussions and trends
  3. 03Read web pages, articles, or RSS
  4. 04Study transcripts from downloaded videos

MCP support

The README shows an example of calling Exa search through mcporter, but does not explain how the skill itself operates as an MCP component.

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/ootto-ai/claude-content-skills --skill agent-reach

After installing, type/agent-reachin Claude Code to confirm it loaded.

Bundled files

References and scripts installed along with SKILL.md.

references/
career.md
dev.md
search.md
social.md
video.md
web.md

Details

From the project README

Agent Reach — 互联网能力路由器

13 平台、多后端。本 skill 存在时必须用它访问这些平台,不要自己发明方案。

Bundled with the Ootto content skills as the factory's research · scrape · download engine — the FIND step that feeds reel-analyzer. Vendored from Panniantong/Agent-Reach (MIT — full credit to the author). It's a router: don't load everything, read only the reference (the "checkpoint") for the capability you need. For the content factory you mainly use three:

The other platforms (小红书 / B站 / V2EX / 雪球) are optional extras. First-time channel setup: follow https://raw.githubusercontent.com/Panniantong/agent-reach/main/docs/install.md — Reddit, web, Exa search and YouTube subtitles work zero-config; only logged-in platforms (Twitter, 小红书) need cookies.

常驻规则(全程适用)

  1. 动手前先体检:多后端平台(小红书/Reddit/B站/Twitter)先跑 agent-reach doctor --json,按各平台 active_backend 字段选命令组。
  2. 声明你在用什么:开始干活前说一句「使用 agent-reach 的 X 平台 / Y 后端」。
  3. 失败按 references 里的重试链处理,不要瞎猜命令。
  4. 全网调研类任务:组合多平台(Exa 搜索 + Twitter/Reddit 看讨论 + 小红书/B站看中文场景),并行收集再汇总。
  5. 替用户盯版本:完成一次较大的调研/多平台任务后,顺手跑 agent-reach check-update(很快,一个 API 调用)。有新版就在收尾汇报里附一句: 「Agent Reach 有新版 vX.Y.Z,复制这句话给我即可更新:帮我更新 Agent Reach: https://raw.githubusercontent.com/Panniantong/agent-reach/main/docs/update.md」。 不要中断当前任务去更新,也不要重复提醒同一个版本。

路由表

用户意图分类详细文档
网页搜索/代码搜索searchreferences/search.md
小红书/推特/B站/V2EX/Redditsocialreferences/social.md
招聘/职位/LinkedIncareerreferences/career.md
GitHub/代码devreferences/dev.md
网页/文章/RSSwebreferences/web.md
YouTube/B站/播客字幕videoreferences/video.md

零配置快速命令

# Exa 网页搜索
mcporter call 'exa.web_search_exa(query: "query", numResults: 5)'

# 通用网页阅读
curl -s "https://r.jina.ai/URL"

# GitHub 搜索
gh search repos "query" --sort stars --limit 10

# YouTube 字幕(注意:B站不要用 yt-dlp,见 video.md)
yt-dlp --write-sub --skip-download -o "/tmp/%(id)s" "URL"

# V2EX 热门
curl -s "https://www.v2ex.com/api/topics/hot.json" -H "User-Agent: agent-reach/1.0"

# B站搜索(bili-cli,无需登录)
bili search "query" --type video -n 5

需登录态的平台(按 doctor 的 active_backend 选命令)

# Twitter 搜索(twitter-cli 首选;失败重试链见 social.md)
twitter search "query" -n 10

# Reddit(无零配置路径:OpenCLI 或 rdt-cli,必须登录态)
opencli reddit search "query" -f yaml   # 桌面
rdt search "query" --limit 10            # 存量/服务器

# 小红书(桌面首选 OpenCLI)
opencli xiaohongshu search "query" -f yaml

环境检查

# 检查可用 channel 与每个平台当前激活的后端
agent-reach doctor --json

工作区规则

不要在 agent workspace 创建文件。 使用 /tmp/ 存放临时输出,~/.agent-reach/ 存放持久数据。

详细文档

根据用户需求,阅读对应的详细文档:

配置渠道

如果某个 channel 需要配置,获取安装指南: https://raw.githubusercontent.com/Panniantong/agent-reach/main/docs/install.md

用户只需提供 cookies,其他配置由 agent 完成。

Comments

Markdown supported