ginobefun/bestblogs · Archived

bestblogs-read

Use this skill when the user wants to deep-read a specific BestBlogs resource (article / podcast / video / tweet) — fetch its metadata + markdown so a local LLM or reader can consume it, and automatically report the read behavior back to BestBlogs. Triggers include "深度?

First seen May 19, 2026

Installation

$ npx skills add ginobefun/bestblogs --skill bestblogs-read

Stronger alternatives

This repository is archived — consider an actively maintained alternative.

Also in this package

Other skills from ginobefun/bestblogs.

npx skills add ginobefun/bestblogs

Browse all from ginobefun/bestblogs

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Repository health

Stars 4.0K
Default branch main
Open issues 12
Status Archived

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 2,868 B
  • docs SUMMARY.md 494 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 5 installs

SKILL.md

bestblogs-read — 资源深度阅读(Local)

何时使用

  • 用户想让本地 LLM / reading UI 消费某篇 BestBlogs 内容的正文。
  • 用户想看某条资源的元信息 + 全文 Markdown。
  • 用户想在阅读的同时把 "read" 行为回传,让 BestBlogs 的个性化飞轮跟得上。

原语

原语 CLI 说明
bestblogs.resource.meta.get 内部调用 /resources/{id}/meta meta
bestblogs.resource.markdown.get 内部调用 /resources/{id}/markdown 正文 markdown
bestblogs.resource.read.mark 内部调用 POST /resources/{id}/read 回传已读
bestblogs.resource.deep_read.prepare bestblogs read deep <id> 一次完成 meta + markdown + mark read

只暴露 deep_read.prepare 作为主原语——它是 CLI bestblogs read deep 的 wrap,并行拉 meta+markdown,串行 mark read。

动作

完整深读(默认,包含 mark read)

bestblogs read deep RAW_xxx --json

返回:

{
  "success": true,
  "data": {
    "meta": {
      "resourceId": "RAW_xxx",
      "title": "...",
      "url": "...",
      "resourceType": "ARTICLE",
      "language": "en_US",
      "score": 88,
      "sourceName": "..."
    },
    "markdown": "# Title\n\n...",
    "readMarked": true,
    "markError": null
  }
}

只要正文 Markdown(喂给本地 LLM)

bestblogs read deep RAW_xxx --markdown-only

输出纯 Markdown(stdout),适合 | anthropic / | llm 等管道。

不回传 read 行为(如只是预览)

bestblogs read deep RAW_xxx --no-mark-read --json

错误处理

  • 404 / NOT_FOUND → 资源不存在或已下架;告诉用户
  • 资源为私有源(sourceVisibility=PRIVATE)且当前用户未订阅 → meta 可能返回 null;提示用户订阅对应源
  • mark read 失败时 CLI 返回 readMarked=false, markError="..." 而不是整体失败,skill 应保留正文给用户并告知 "已读回传失败,不影响阅读"

组合建议

  • 常用链:bestblogs-discover today → 从 candidates[] 挑一条 → bestblogs-read deep <id> --json → 把 markdown 喂给本地 LLM 做总结/提问 → bestblogs-capture 做笔记保存
  • 不要重复回传:默认就会 mark read,除非用户明确只是预览再决定是否读