Save My Disk — Security skill for Claude Code
Universal AI Agent skill & runbook for intelligent, non-invasive disk audit and safe optimization.
How to install Save My Disk
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open Sorakis/save-my-disk and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Save My Disk does
Universal AI Agent skill & runbook for intelligent, non-invasive disk audit and safe optimization. Supports Antigravity, WorkBuddy, Cursor, Claude Code, and Codex.
Alternatives in Security
- Anthropic Cybersecurity Skills — 734+ structured cybersecurity skills for AI agents · MITRE ATT&CK mapped · agentskills.io open standard · Work 3.8k ★
- Audit Agent Sessions — Analyze all agent sessions from the last 3 days across Claude, Codex, and Gemini 238 ★
- Loongsuite Pilot — Local-first telemetry collector for AI coding agents — unified OpenTelemetry events for Claude Code, Codex, Cu 150 ★
README
Save My Disk 🚀 (save-my-disk)
**Universal AI Agent Skill & Runbook for Intelligent, Non-Invasive Disk Optimization.** 面向 AI 编程助手(Antigravity, OpenAI Codex, Claude Code, WorkBuddy, OpenCode, DeepSeek Harness, Cursor 等)与广大开发者的智能磁盘诊断与安全瘦身工作流。
[](LICENSE) [](https://github.com/Sorakis/save-my-disk/actions) [](https://github.com/Sorakis/save-my-disk/releases) [%20%7C%20Linux%20%26%20macOS%20(Planned)-0078D6.svg)]() []()
🌟 核心理念 (Core Philosophy)
- 只读优先 (Read-Only First):初筛体检阶段严禁执行任何写操作,严格规避死循环与高 I/O 锁盘,前台进程丝滑不卡顿。
- 四级风险决策矩阵 (Risk-Tiered Triage):
- 🟢 Tier 1 (无痛零风险):系统更新补丁、
$WINDOWS.~BT、临时文件、传递优化。 - 🟡 Tier 2 (开发者缓存 / 可迁移项):Docker 虚拟磁盘、Conda/pip/npm 缓存、AI 模型临时运行库。
- 🟠 Tier 3 (应用内治理与残留):Steam 卸载空残留、办公/聊天日志转储、微信与飞书媒体数据。
- 🔴 Tier 4 (绝对红线禁区):
System32、WinSxS、正在工作的项目源码。
- 🟢 Tier 1 (无痛零风险):系统更新补丁、
- 双轨闭环执行 (Dual-Track Engine):
- Agent 自动化执行:对零风险补丁缓存,由 Agent 提供一键脚本(必须先获用户明确授权)。
- Agent 人机协作引导:对需要 GUI 权限的操作(如
cleanmgr清理系统文件、Docker 虚拟盘无损迁移到大容量盘),Agent 输出极简分步指引。
- 自愈与战果核验 (Verification):任何清理操作完成后,自动回访核对磁盘剩余空间差值,并验证关键系统服务健康状态。
🏗️ 架构与工作流 (Architecture)
flowchart TD
A["用户提出需求: 盘满了/清理空间"] --> B["Phase 1: 只读轻量体检 (audit_drives.ps1)"]
B --> C["Phase 2: 智能风险分级与决策 (Safety Matrix)"]
C --> D{"清理路径选择"}
D -->|"Tier 1 缓存清理"| E["Track A: Agent 执行脚本 (需用户确认)"]
D -->|"系统升级残留/Docker迁移"| F["Track B: Agent 指导用户执行 (cleanmgr/Docker设置)"]
E --> G["Phase 4: 结果自愈核验 (verify_health.ps1)"]
F --> G
G --> H["输出战果报表: 释放空间量 + 服务健康确认"]
🤖 多 Agent 支持规范 (
Related Skills
HyperArmor Skill
🛡️ HyperArmor-Skill: 商业级全自动二进制深度加固与防逆向 Agent Skill (Universal Security Skill for Antigravity, Cursor, Claude
Aegis AI Agent Security
Enterprise AI-coding-agent security: a lightweight endpoint agent governs Cursor, Claude Code, Codex, Windsurf
Production Agent Skills
High-reliability, deterministic Agent Skills for autonomous AI coding agents (Antigravity, Claude Code, Cursor
Claude Code Bash Guardian
Automated security layer for Claude Code PreToolUse Bash Tool hooks. Provides bash command filtering, bypass p
Millieskills
Research-driven universal AI Agent Skills for UI/UX, code repair, security, testing, architecture, backend/API
Universal Re Skill
Universal reverse engineering, 40+ vuln classes & working PoC generation for Android/iOS/Windows PE/Linux ELF/
Related Agents
Model Runner
Deterministic wrapper that runs a prompt on a non-Claude model (OpenAI via Codex, Cursor's composer/grok/glm,
Paf Reviewer
Runs one phone-a-friend relay or review command in the background and reports back a receipt plus the findings
Security Sentinel
Deep security review focused on OWASP Top 10, injection, auth/authz, data exposure, and dependency vulnerabili