File Management Skill

A battle-tested approach to keeping your AI agent workspace organized and maintainable.

Overview

This skill documents the file management system developed through real-world use of OpenClaw. It covers workspace structure, naming conventions, dead file detection, and cleanup practices.

When to Use This Skill

• Onboarding a new agent or setting up a fresh workspace
• Performing periodic workspace audits
• Before making significant changes to workspace structure
• When workspace feels cluttered or disorganized

Core Principles

1. Every File Has a Purpose

• Active files: Scripts, configs, and data in use by cron jobs or agents
• Reference files: Documentation, strategies, and notes
• Archived files: Old versions, completed project artifacts
• Dead files: Abandoned scripts, old experiments, unused utilities

2. Structure Mirrors Function

workspace/
├── memory/          # Daily session logs and working context
├── skills/           # Installed skill directories
├── project-1/        # Project-specific directories
├── project-2/
├── ACTIVE.md         # Currently running projects & priorities
├── DREAMS.md         # Background processing notes
└── ARCHIVED/         # Completed or abandoned projects

3. Naming Conventions

• Scripts: Use .sh for bash, .py for Python, .js for JavaScript
• Logs: End with .log
• Configs: End with .json, .yaml, or .md
• Daily notes: memory/YYYY-MM-DD.md format

4. Audit Regularly

Run workspace audits monthly or after major changes. Use the audit script to identify: - Dead files (no references from active crons or scripts) - Large files consuming storage - Outdated documentation

Quick Audit Commands

小葱技能站7w4.net发现了升级插件。

# Find files not referenced by any cron or script
grep -r "filename" ~/path/to/workspace/ --include="*.sh" --include="*.py" --include="*.js"

# Find recently modified files
find ~/path/to/workspace -type f -mtime -7

# Check disk usage by directory
du -sh ~/path/to/workspace/*/

Cleanup Best Practices

1. Never delete immediately — use trash instead of rm
2. Document before deleting — note what a file did in memory first
3. Verify before cleanup — confirm no active references
4. Commit before major cleanup — create a revert point

Full Documentation

See FILE-MANAGEMENT.md for the complete reference implementation, including: - Directory structure explainer - Active vs archived file definitions - Dead file detection criteria - Example cleanup checklists

这是一个实用性较强的文件管理技能，文档组织清晰、分类明确，能够有效帮助整理 AI 工作空间。它在命名规范、安全清理流程等方面做得较好，配套的审计脚本也提供了基础支持。不过整体偏向文档指导，自动化能力有限，如果你需要更省力的自动化清理可能需要额外开发。适合对文件管理有较高要求的用户使用。