Token导航 LogoToken导航TokenDH.com
研究检索执行命令clawhub未标认证来源可访问clear审计提醒

scribescribe 搜索

Agent Skill

scribe 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。

总安装

12,920

周安装

533

GitHub Stars

公开资料未说明

下载量

4,221
OpenClaw

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT-0

最后核验

2026-05-01

来源状态

来源可访问

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

简介

scribe 用于扫描 OpenClaw 日志和行为数据,生成每日和每周的笔记与总结。

  • 适用于行为分析、习惯追踪和工作复盘等效率提升场景。
  • 通过 clawhub 安装并使用 openclaw skills install scribe 命令部署。
  • 使用前需确认权限范围、维护状态,注意可能触发联网、命令执行或文件读写操作。
  • 建议结合来源仓库和原始 README 文档进一步核验具体用法和功能边界。

SKILL.md

name
scribe
displayName
Scribe | OpenClaw Skill
description
Scans OpenClaw logs, config files, chat history, cursor history, behavior, desires, tastes, and drafts to take comprehensive daily and weekly notes with summaries.
version
1.0.0

Scribe | OpenClaw Skill

Comprehensive note-taking system that scans multiple OpenClaw data sources to generate daily and weekly notes with summaries.

Description

Scribe automatically scans and analyzes:

  • OpenClaw logs - Errors, warnings, gateway events, subagent activity
  • Config files - Model preferences, gateway settings, agent configurations
  • Chat history - Cursor IDE chat messages from SQLite databases
  • Cursor history - Workspace-specific conversation history
  • Memory files - Daily notes and long-term memory (MEMORY.md)
  • Behavior files - BEHAVIOR.md, DESIRES.md, TASTES.md, PREFERENCES.md
  • Drafts - Blog posts, tweets, and other draft files

Generates structured markdown notes with summaries at the top of each file.

Installation

clawhub install scribe

Or clone into your skills directory:

git clone https://github.com/Org/scribe.git workspace/skills/scribe

Usage

Daily Notes

Generate a daily note covering the last 24 hours:

python3 workspace/skills/scribe/scripts/scribe.py --mode daily

Weekly Notes

Generate a weekly note covering the last 7 days:

python3 workspace/skills/scribe/scripts/scribe.py --mode weekly

Both Daily and Weekly

Generate both notes at once:

python3 workspace/skills/scribe/scripts/scribe.py --mode both

JSON Output

Get results in JSON format:

python3 workspace/skills/scribe/scripts/scribe.py --mode daily --json

Examples

Example 1: Daily Note Generation *Scenario:* You want a daily summary of OpenClaw activity. *Action:* Run python3 workspace/skills/scribe/scripts/scribe.py --mode daily. *Outcome:* A markdown file workspace/Notes/daily/YYYY-MM-DD.md with a summary at the top, followed by detailed sections for logs, chat history, memory, drafts, behavior, and config.

Example 2: Weekly Summary *Scenario:* You want a weekly overview of patterns and trends. *Action:* Run python3 workspace/skills/scribe/scripts/scribe.py --mode weekly. *Outcome:* A markdown file workspace/Notes/weekly/YYYY-MM-DD.md with weekly statistics, trends, and activity summaries.

Example 3: Cron Job Integration *Scenario:* Automate daily note generation. *Action:* Add a cron job that runs scribe.py --mode daily every day at midnight. *Outcome:* Daily notes are automatically generated and saved to workspace/Notes/daily/.

Commands

python3 workspace/skills/scribe/scripts/scribe.py --mode daily    # Generate daily note
python3 workspace/skills/scribe/scripts/scribe.py --mode weekly   # Generate weekly note
python3 workspace/skills/scribe/scripts/scribe.py --mode both     # Generate both
python3 workspace/skills/scribe/scripts/scribe.py --mode daily --json  # JSON output
python3 workspace/skills/scribe/scripts/scribe.py --openclaw-home /path/to/openclaw  # Custom home directory
  • --mode — Choose daily, weekly, or both (default: daily)
  • --json — Output results in JSON format instead of markdown files
  • --openclaw-home — Specify OpenClaw home directory (default: ~/.openclaw)

What this skill does

  1. Scans logs - Analyzes all .log files in logs/ directory for errors, warnings, gateway events, and subagent spawns
  2. Reads config - Extracts model preferences, gateway settings, and agent configurations from openclaw.json
  3. Extracts chat history - Queries Cursor's SQLite databases (state.vscdb) for recent chat messages
  4. Scans memory files - Reads daily notes (memory/YYYY-MM-DD.md) and long-term memory (MEMORY.md)
  5. Finds drafts - Searches for draft files matching patterns like *draft*.txt, *draft*.md, blog/**/*.md, tweet*.txt
  6. Reads behavior files - Looks for BEHAVIOR.md, DESIRES.md, TASTES.md, PREFERENCES.md
  7. Generates notes - Creates structured markdown files with summaries at the top
  8. Saves to Notes - Writes daily notes to workspace/Notes/daily/YYYY-MM-DD.md and weekly notes to workspace/Notes/weekly/YYYY-MM-DD.md

Output Format

Daily Note Structure

# Daily Summary - YYYY-MM-DD HH:MM:SS

- **Logs**: X errors, Y warnings
- **Gateway Events**: Z events
- **Subagent Activity**: N spawns
- **Chat Messages**: M messages
- **Daily Notes**: K files
- **Drafts**: L draft files
- **Config**: Loaded successfully

---

# Daily Note

## Logs
### Errors (X)
- [error details]

### Warnings (Y)
- [warning details]

## Chat History
Found M messages in the last 24 hours.
### User (X messages)
- [message previews]

## Memory Files
### Daily Notes (K)
- **YYYY-MM-DD**: [content preview]

## Drafts
### [draft path]
- Size: X bytes
- Modified: [timestamp]
- Preview: [content preview]

## Behavior & Preferences
### Behavior Files
- **BEHAVIOR.md**: [content preview]

## Configuration
### Model Preferences
- Default: [model]
- Aliases: X configured

Weekly Note Structure

# Weekly Summary - YYYY-MM-DD HH:MM:SS

[Summary statistics]

---

# Weekly Note - YYYY-MM-DD to YYYY-MM-DD

## Weekly Summary
- **Total Errors**: X
- **Total Warnings**: Y
- **Gateway Events**: Z
- **Subagent Spawns**: N

## Chat Activity (M messages)
- **User**: X messages
- **Assistant**: Y messages

## Memory Activity (K daily notes)
- **YYYY-MM-DD**: X bytes

## Drafts (L files)
- **[path]**: X bytes (modified: [timestamp])

## Trends & Patterns
- ⚠️ **Error Rate**: X errors this week
- 🤖 **Subagent Activity**: N spawns this week
- 💬 **Chat Activity**: M messages this week

Requirements

  • Python 3.7+
  • OpenClaw installation with ~/.openclaw directory structure
  • Cursor IDE installed (for chat history scanning)
  • SQLite3 available (usually pre-installed on macOS)
  • Write access to workspace/Notes/ directory

Integration as a Cron Job

Example Cron Job Configuration:

{
  "payload": {
    "kind": "agentTurn",
    "message": "Run scribe.py --mode daily to generate daily notes.",
    "model": "openrouter/google/gemini-2.5-flash",
    "thinking": "low",
    "timeoutSeconds": 300
  },
  "schedule": {
    "kind": "cron",
    "cron": "0 0 * * *"
  },
  "delivery": {
    "mode": "announce"
  },
  "sessionTarget": "isolated",
  "name": "Daily Scribe Note"
}

Or run directly via shell script:

# Add to crontab (crontab -e)
# Run daily at midnight
0 0 * * * /Users/ghost/.openclaw/workspace/skills/scribe/scripts/scribe.py --mode daily >> /Users/ghost/.openclaw/logs/scribe.log 2>&1

Security & Privacy

  • File access: Reads openclaw.json (config only, no secrets), logs/*.log, memory/*.md, and Cursor SQLite databases
  • No data exfiltration: All data stays local; notes are saved to workspace/Notes/ directory
  • Safe execution: Uses read-only access to config and logs; only writes to workspace/Notes/daily/ and workspace/Notes/weekly/ directories

Limitations

  • Chat history extraction depends on Cursor's database schema (may need updates if Cursor changes storage format)
  • Draft file detection uses pattern matching; may miss files with non-standard naming
  • Behavior file detection looks for common filenames; custom locations may not be found
  • Large log files may take time to process

适合场景

01

调用多模型

02

代码和文本生成

03

Agent 推理流程

04

OpenRouter 模型接入

能力概览

能力 1

统一调用多种 LLM

能力 2

支持 Claude、Gemini、Kimi 等模型

能力 3

适合聊天、代码和推理任务

能力 4

可作为 Agent 模型调用入口

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenClaw

98.23%
按下载量换算4,146

安全审计

VirusTotal

可疑

ClawScan

可疑

Static analysis

未展示

权限和风险

执行命令

安装流程涉及命令执行,可能通过 openclaw skills install scribe 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

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 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills