Token导航 LogoToken导航

Agent Skill · 研究检索

logseq(logseq 搜索)

logseq 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。

logseq 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为操作浏览器;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 1,183。

总安装 112,694周安装 4,515GitHub Stars 6下载 36,481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
1.0.0

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:logseq(logseq 搜索)
来源仓库:https://github.com/juanirm/logseq
安装命令:
openclaw skills install logseq
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 logseq

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。

提供用于通过其插件 API 与本地 Logseq 实例交互的命令。用于在 Logseq 中创建页面、插入块、查询图形数据库、管理任务、检索内容或自动化工作流程。仅适用于启用了 API 的本地运行实例; [$API 可访问技能] 所需的默认端口或设置路径。

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 logseq;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

    它需要操作浏览器,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

logseq 怎么安装?

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

logseq 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。

logseq 需要什么权限?

权限级别是操作浏览器。该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

logseq 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。

logseq 用的人多吗?

公开数据是累计安装 112,694 次,近一周安装 4,515 次,下载 36,481 次,GitHub 6 星。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。

logseq 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
logseq

logseq 搜索 · 本页

112,6946操作浏览器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
logseq
description
Provide commands for interacting with a local Logseq instance through its Plugin API. Use for creating pages, inserting blocks, querying the graph database, managing tasks, retrieving content, or automating workflows in Logseq. Only works with a locally running instance with the API enabled; default port or set path expected for [$API accessible skill].

Logseq Plugin API

Interact with your local Logseq instance through its JavaScript Plugin API. This skill enables reading, writing, querying, and automating workflows in your Logseq graph.

Prerequisites

Logseq must be running locally with a plugin that exposes the API. The standard way is:

  1. Install a bridge plugin that exposes logseq API via HTTP (e.g., via a custom plugin or localhost endpoint)
  2. Alternative: Use Node.js with @logseq/libs package to script against the running Logseq instance

The API is primarily designed for in-browser plugins, so accessing it from external scripts requires a bridge/proxy.

Core API Namespaces

The Logseq Plugin API is organized into these main proxies:

logseq.App

Application-level operations: getting app info, user configs, current graph, commands, UI state, external links.

Key methods:

  • getInfo() - Get app version and info
  • getUserConfigs() - Get user preferences (theme, format, language, etc.)
  • getCurrentGraph() - Get current graph info (name, path, URL)
  • registerCommand(type, opts, action) - Register custom commands
  • pushState(route, params, query) - Navigate to routes

logseq.Editor

Block and page editing operations: creating, updating, moving, querying content.

Key methods:

  • getBlock(uuid) - Get block by UUID
  • getCurrentPage() - Get current page entity
  • getCurrentPageBlocksTree() - Get all blocks on current page
  • getPageBlocksTree(page) - Get all blocks for a specific page
  • insertBlock(target, content, opts) - Insert a new block
  • updateBlock(uuid, content) - Update block content
  • createPage(pageName, properties, opts) - Create a new page
  • deletePage(pageName) - Delete a page
  • getPageLinkedReferences(page) - Get backlinks to a page
  • registerSlashCommand(tag, action) - Add custom slash commands

logseq.DB

Database queries using Datalog.

Key methods:

  • q(query, ...inputs) - Run Datalog query
  • datascriptQuery(query, ...inputs) - Direct Datascript query

logseq.UI

UI operations: messages, dialogs, main UI visibility.

Key methods:

  • showMsg(content, status) - Show toast notification
  • queryElementById(id) - Query DOM elements

logseq.Git

Git operations for the current graph.

Key methods:

  • execCommand(args) - Execute git command

logseq.Assets

Asset management.

Key methods:

  • listFilesOfCurrentGraph(path) - List files in graph

Common Workflows

Read Content

// Get current page
const page = await logseq.Editor.getCurrentPage();

// Get all blocks on a page
const blocks = await logseq.Editor.getPageBlocksTree('Daily Notes');

// Get a specific block
const block = await logseq.Editor.getBlock('block-uuid-here');

// Query with Datalog
const results = await logseq.DB.q(`
  [:find (pull ?b [*])
   :where [?b :block/marker "TODO"]]
`);

Write Content

// Create a new page
await logseq.Editor.createPage('Project Notes', {
  tags: 'project',
  status: 'active'
}, { redirect: false });

// Insert a block
const block = await logseq.Editor.insertBlock(
  'target-block-uuid',
  '- New task item',
  { before: false, sibling: true }
);

// Update a block
await logseq.Editor.updateBlock('block-uuid', 'Updated content');

// Batch insert multiple blocks
const blocks = [
  { content: 'First item' },
  { content: 'Second item', children: [
    { content: 'Nested item' }
  ]}
];
await logseq.Editor.insertBatchBlock('parent-uuid', blocks, { sibling: false });

Task Management

// Find all TODO items
const todos = await logseq.DB.q(`
  [:find (pull ?b [*])
   :where
   [?b :block/marker ?marker]
   [(contains? #{"TODO" "DOING"} ?marker)]]
`);

// Mark task as DONE
await logseq.Editor.updateBlock('task-uuid', 'DONE Task content');

// Get tasks on current page
const page = await logseq.Editor.getCurrentPage();
const blocks = await logseq.Editor.getPageBlocksTree(page.name);
const tasks = blocks.filter(b => b.marker === 'TODO' || b.marker === 'DOING');

Navigation and UI

// Navigate to a page
logseq.App.pushState('page', { name: 'Project Notes' });

// Show notification
logseq.UI.showMsg('✅ Task completed!', 'success');

// Get app config
const configs = await logseq.App.getUserConfigs();
console.log('Theme:', configs.preferredThemeMode);
console.log('Format:', configs.preferredFormat);

Implementation Approaches

Since Logseq's Plugin API is browser-based, you have several options:

Option 1: Bridge Plugin

Create a minimal Logseq plugin that exposes API calls via HTTP:

// In Logseq plugin (index.js)
logseq.ready(() => {
  // Expose API endpoints
  logseq.provideModel({
    async handleAPICall({ method, args }) {
      return await logseq.Editor[method](...args);
    }
  });
});

// Then call from external script via HTTP POST

Option 2: Node.js Script with @logseq/libs

For automation scripts, use the @logseq/libs package:

npm install @logseq/libs

Note: This requires a running Logseq instance and proper connection setup.

Option 3: Direct Plugin Development

Develop a full Logseq plugin following the plugin samples at: https://github.com/logseq/logseq-plugin-samples

API Reference

For complete API documentation, see:

  • API Docs: https://logseq.github.io/plugins/
  • Plugin Samples: https://github.com/logseq/logseq-plugin-samples
  • Type Definitions: references/api-types.md (extracted from @logseq/libs)

Key Data Structures

BlockEntity

{
  id: number,           // Entity ID
  uuid: string,         // Block UUID
  content: string,      // Block content
  format: 'markdown' | 'org',
  page: { id: number }, // Parent page
  parent: { id: number }, // Parent block
  left: { id: number }, // Previous sibling
  properties: {},       // Block properties
  marker?: string,      // TODO/DOING/DONE
  children?: []         // Child blocks
}

PageEntity

{
  id: number,
  uuid: string,
  name: string,              // Page name (lowercase)
  originalName: string,       // Original case
  'journal?': boolean,
  properties: {},
  journalDay?: number,       // YYYYMMDD for journals
}

Tips & Best Practices

  1. Always check for null: API methods may return null if entity doesn't exist
  2. Use UUIDs over IDs: Block UUIDs are stable, entity IDs can change
  3. Batch operations: Use insertBatchBlock for multiple inserts
  4. Query efficiently: Datalog queries are powerful but can be slow on large graphs
  5. Properties are objects: Access with block.properties.propertyName
  6. Format matters: Respect user's preferred format (markdown vs org-mode)
  7. Async all the way: All API calls return Promises

Common Gotchas

  • Page names are lowercase: When querying, use lowercase page names
  • Journal pages: Use journalDay format (YYYYMMDD) not date strings
  • Block hierarchy: Respect parent/child relationships when inserting
  • Format differences: Markdown uses - for bullets, Org uses *
  • Properties syntax: Different between markdown (prop::) and org (:PROPERTIES:)

平台分布

OpenClaw

94.84%
按下载量换算34,599

安全审计

VirusTotal

可疑

ClawScan

通过

Static analysis

未展示

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

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