Token导航 LogoToken导航

Agent Skill · 研究检索

pages(pages 搜索)

pages 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

pages 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为需要联网;许可证为 MIT-0,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 24,843。

总安装 3,912周安装 163GitHub Stars 公开资料未说明下载 1,304核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
1.0.1

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:pages(pages 搜索)
来源仓库:https://github.com/ilyabelikin/pages
安装命令:
openclaw skills install pages
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install pages

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

用于管理和检索书籍情报,支持按页数组织 Markdown 文件。适合记录新书、捕捉引言及进行表面阅读等研究任务。每本书对应一个文件,便于结构化存储与快速定位内容。安装后建议检查文件路径权限及读写安全性。适用宿主包括 OpenClaw,接入前应确认版本、权限和运行环境要求。

pages 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install pages;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

pages 怎么安装?

在终端执行 openclaw skills install pages。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

pages 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

pages 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

pages 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

pages 用的人多吗?

公开数据是累计安装 3,912 次,近一周安装 163 次,下载 1,304 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

pages 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。

先想清楚的地方

  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
pages

pages 搜索 · 本页

3,912—需要联网MIT-0有提醒

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
pages
description
Reading intelligence — log books, capture quotes, and surface reads when they're relevant. One markdown file per book in pages/. Use when logging a new book, searching for what you've read, or asking "what have I read about X?
metadata
openclaw
emoji
📖
os
["linux", "darwin", "win32"]
hermes
tags
["reading", "books", "knowledge"]

📖 Pages — reading intelligence

Data

Base path is workspace root or document root folder. On first use, create it: mkdir -p mind/pages/. Pages uses a mind/pages/ folder in your workspace.

Files live in mind/pages/. Organised by reading status:

mind/
└── pages/
    ├── pagesconfig.yml
    ├── read/        ← finished
    ├── reading/     ← currently reading
    └── want/        ← want to read

File names: <author-lastname>-<title-slug>.md. Example: taleb-antifragile.md. For books with no clear author (anthologies, edited volumes): use the title slug.


Dataset Config — pagesconfig.yml

pagesconfig.yml lives inside the mind/pages/ directory. Read it at the start of any session involving this skill.

images: no (by default no, ask if you human want to feach images for pepople, warn that it is token expensive)

Book File

# Book Title

Author: Full Name
Year: 2012
Finished: 12 Feb 2026 (omit if not finished)
Rating: 5/5 (omit until finished)
Image: if `mind/pages/pagesconfig.yml` has `images: yes` add a link to an image in `../assets/slug-that-make-sense`
Tags: #philosophy #risk #systems #investing
Recommended by: [[marco-tabini]] (Peeps slug — omit if unknown)
Goodreads: link to goodreads

## Notes

12 Feb 2026: key insight, what changed your mind, what you'd tell someone else

## Quotes

> "The wind extinguishes a candle and energizes fire."
> — p. 61

Field guidance:

Tags: 2–4 themes, domains, or moods. Use tags you'll actually search: #ai not #artificial-intelligence, #startups not #entrepreneurship. Keep it personal. Recommended by: if Peeps is installed, use [[their-slug]]; otherwise note the person's name as plain text. Creates a bidirectional trail when Peeps is present. Notes: dated insight log, not a summary. What this book changed, challenged, or confirmed. "Good book" is not a note. Quotes: only the lines worth keeping. One good quote beats ten mediocre ones.

The want/ folder: lightweight. Just title, author, and a one-line Why: note on why it's on your list. Add more once you start reading.


Saving a Book

  1. Search the web (title + author) — confirm full author name, publication year.
  2. Determine folder — which folder to save to: read/, reading/, or want/.
  3. Ask as a group (skip anything already clear):

- If read: rating and finished date? - Tags — what's this book about, in your words? - Any quotes or notes to capture right now? - Who recommended it? 4. if images: yes in mind/pagesconfig.yml search for a conver image and add it to Image: feild

Show a brief confirmation: "Saved — *Antifragile* by Nassim Taleb, 2012, in read/. Tagged #philosophy #risk #systems."


Core Behavior

  • User mentions a book → check if saved, offer to create or update
  • User asks "what have I read about X?" → search mind/pages/ with expanded keywords
  • User finishes a book → ask for a rating and a note
  • User mentions someone recommending a book → save with Recommended by:; if Peeps is installed, optionally add a note to their Peeps file
  • User shares a quote → append to the book file with today's date
  • Conversation touches a theme → surface relevant reads contextually without being asked

Examples:

  • "Just finished Thinking, Fast and Slow" → check if saved, offer to rate and note
  • "I need to think about decision-making under pressure" → "You read *Thinking, Fast and Slow* in 2024 and tagged it #decision-making — your note says it reframed how you think about bias"
  • "Priya said I should read *The Mom Test*" → save to mind/pages/want/ with Recommended by: [[priya-nair]] if Peeps is installed, otherwise note her name as plain text

Finding Books

Use grep with expanded terms. Always broaden the query before searching.

# Books on a topic
grep -ril "decision\|bias\|judgement\|psychology" mind/pages/

# Highly-rated reads
grep -rl "Rating: 5" mind/pages/read/

# Books recommended by a specific person
grep -rl "\[\[priya" mind/pages/

# Currently reading
ls mind/pages/reading/

# Books with captured quotes
grep -rl "^>" mind/pages/

Keyword expansion examples:

  • "business" → business\|strategy\|management\|leadership\|startup\|founder
  • "productivity" → productivity\|focus\|habits\|deep.work\|flow\|systems
  • "tech / AI" → ai\|machine.learning\|software\|internet\|tech\|silicon
  • "philosophy" → philosophy\|stoic\|ethics\|meaning\|existence\|wisdom

Always read the full book file after grepping — matched tags are a signal, not the full picture.


Heartbeat or cron

Check a random book file. Surface something worth mentioning:

  • "You finished *Antifragile* three months ago and never rated it — worth a quick reflection?"
  • "You added *The Mom Test* to your want list in January — still on your radar?"
  • "Priya and you have both tagged #systems — you might be reading similar things."

If nothing worth mentioning, skip.


Adding to HEARTBEAT.md or cron

If it is not there yet, ask your human if they want to add Pages: check to HEARTBEAT.md. If there is no HEARTBEAT.md, suggest to create a cron every 30 minutes during waking hours (*/30 7-22 * * *) to execute Pages: check.


Config — pagesconfig.yml

Optional. Lives inside mind/pages/.

owner: jane-smith  # your Peeps slug — used for cross-referencing shared reads

If Peeps is installed and peepsconfig.yml exists, owner is inferred automatically — no need to duplicate.


Integration with Peeps

If Peeps is installed, books and people can be connected:

Recommended by: [[their-slug]] in the book file — who pointed you at this

  • Optionally add a note in their Peeps file: "Recommended *Book Title* (Mar 2026)"
  • When the conversation involves a person and a relevant read comes up, surface it: "You and Marco both read *Zero to One* — his note is that Thiel's distribution thinking changed how he pitches."

Integration with Haah

If Haah is installed, dispatch to your circles when your reading list is thin on a topic:

  • "Haah: anyone in my circles with strong reads on climate tech?"

When someone in your circle asks for book recommendations:

  • Check Pages for relevant reads before answering
  • Draft a reply with your actual read + rating + one sentence of context. Don't recommend books you haven't read.

Updating

To update this skill to the latest version, fetch the new SKILL.md from GitHub and replace this file:

https://raw.githubusercontent.com/haah-ing/pages-skill/main/SKILL.md

What NOT to Suggest

  • Syncing with Goodreads or Amazon — different purpose, keep private
  • Automatically importing your entire reading history — build it naturally over time
  • Rating every book — only rate when you have something to say
  • Summaries or book reports — this is a memory layer, not a study tool
  • Adding books to others' want lists on their behalf — your list, your choices

平台分布

OpenClaw

78.7%
按下载量换算1,026

安全审计

VirusTotal

可疑

ClawScan

通过

Static analysis

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills