compress

Compress natural language memory files (CLAUDE.md, todos, preferences) into caveman format to save input tokens. Preserves all technical substance, code, URLs, and structure. Compressed version overwrites the original file. Human-readable backup saved as FILE.original.md. Trigger: /caveman:compress <filepath> or "compress memory file"

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "compress" with this command: npx skills add juliusbrussee/caveman/juliusbrussee-caveman-caveman-compress

Caveman Compress

Purpose

Compress natural language files (CLAUDE.md, todos, preferences) into caveman-speak to reduce input tokens. Compressed version overwrites original. Human-readable backup saved as <filename>.original.md.

Trigger

/caveman:compress <filepath> or when user asks to compress a memory file.

Process

  1. This SKILL.md lives alongside scripts/ in the same directory. Find that directory.

  2. Run:

cd <directory_containing_this_SKILL.md> && python3 -m scripts <absolute_filepath>

  1. The CLI will:
  • detect file type (no tokens)
  • call Claude to compress
  • validate output (no tokens)
  • if errors: cherry-pick fix with Claude (targeted fixes only, no recompression)
  • retry up to 2 times
  1. Return result to user

Compression Rules

Remove

  • Articles: a, an, the
  • Filler: just, really, basically, actually, simply, essentially, generally
  • Pleasantries: "sure", "certainly", "of course", "happy to", "I'd recommend"
  • Hedging: "it might be worth", "you could consider", "it would be good to"
  • Redundant phrasing: "in order to" → "to", "make sure to" → "ensure", "the reason is because" → "because"
  • Connective fluff: "however", "furthermore", "additionally", "in addition"

Preserve EXACTLY (never modify)

  • Code blocks (fenced ``` and indented)
  • Inline code (backtick content)
  • URLs and links (full URLs, markdown links)
  • File paths (/src/components/..., ./config.yaml)
  • Commands (npm install, git commit, docker build)
  • Technical terms (library names, API names, protocols, algorithms)
  • Proper nouns (project names, people, companies)
  • Dates, version numbers, numeric values
  • Environment variables ($HOME, NODE_ENV)

Preserve Structure

  • All markdown headings (keep exact heading text, compress body below)
  • Bullet point hierarchy (keep nesting level)
  • Numbered lists (keep numbering)
  • Tables (compress cell text, keep structure)
  • Frontmatter/YAML headers in markdown files

Compress

  • Use short synonyms: "big" not "extensive", "fix" not "implement a solution for", "use" not "utilize"
  • Fragments OK: "Run tests before commit" not "You should always run tests before committing"
  • Drop "you should", "make sure to", "remember to" — just state the action
  • Merge redundant bullets that say the same thing differently
  • Keep one example where multiple examples show the same pattern

CRITICAL RULE: Anything inside ... must be copied EXACTLY. Do not:

  • remove comments
  • remove spacing
  • reorder lines
  • shorten commands
  • simplify anything

Inline code (...) must be preserved EXACTLY. Do not modify anything inside backticks.

If file contains code blocks:

  • Treat code blocks as read-only regions
  • Only compress text outside them
  • Do not merge sections around code

Pattern

Original:

You should always make sure to run the test suite before pushing any changes to the main branch. This is important because it helps catch bugs early and prevents broken builds from being deployed to production.

Compressed:

Run tests before push to main. Catch bugs early, prevent broken prod deploys.

Original:

The application uses a microservices architecture with the following components. The API gateway handles all incoming requests and routes them to the appropriate service. The authentication service is responsible for managing user sessions and JWT tokens.

Compressed:

Microservices architecture. API gateway route all requests to services. Auth service manage user sessions + JWT tokens.

Boundaries

  • ONLY compress natural language files (.md, .txt, extensionless)
  • NEVER modify: .py, .js, .ts, .json, .yaml, .yml, .toml, .env, .lock, .css, .html, .xml, .sql, .sh
  • If file has mixed content (prose + code), compress ONLY the prose sections
  • If unsure whether something is code or prose, leave it unchanged
  • Original file is backed up as FILE.original.md before overwriting
  • Never compress FILE.original.md (skip it)

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

General

caveman

No summary provided by upstream source.

Repository SourceNeeds Review
Coding

CodeBuddy Coding

# CodeBuddy Coding Skill **版本:** 1.0.0 **创建时间:** 2026-03-31 **作者:** OpenClaw Team --- ## 📋 Skill 概述 **CodeBuddy Coding Skill** 是一个通用的 AI 编程能力扩展,让任何 OpenClaw agent 都能调用 CodeBuddy CLI 的强大功能。 ### 核心能力 - ✅ **AI 编程** - 代码生成、重构、调试、优化 - ✅ **文件操作** - 创建、修改、删除文件 - ✅ **命令执行** - 运行构建、测试、部署命令 - ✅ **进度监控** - 实时报告任务进度和状态 - ✅ **结构化输出** - JSON 格式的可解析输出 ### 适用场景 - **Developer Agent** - 编写代码、修复 Bug - **Architect Agent** - 生成项目脚手架 - **Tester Agent** - 编写测试用例 - **任何需要编程能力的 Agent** - 通用编程支持 --- ## 🚀 快速开始 ### 基本用法 ```javascript // 1. 加载 Skill const codebuddy = require('./skill'); // 2. 执行编程任务 const result = await codebuddy.execute({ task: '创建一个用户登录页面', context: { projectPath: '/path/to/project', techStack: 'Vue 3 + TypeScript' }, options: { outputFormat: 'json', permissionMode: 'bypassPermissions' } }); // 3. 获取结果 console.log(result.status); // 'success' console.log(result.filesModified); // ['src/views/Login.vue'] console.log(result.toolCalls); // [{tool: 'write_to_file', ...}] ``` ### 监听进度 ```javascript // 订阅进度事件 codebuddy.onProgress((progress) => { console.log(`进度: ${progress.percentage}%`); console.log(`当前任务: ${progress.currentTask}`); console.log(`已用时间: ${progress.elapsedTime}s`); }); ``` --- ## 🔧 配置说明 ### 环境要求 - **CodeBuddy CLI** v2.68.0+ - **Node.js** v16.0.0+ - **OpenClaw** coding-agent skill 框架 ### Skill 配置 ```json { "name": "codebuddy-coding", "version": "1.0.0", "type": "coding", "capabilities": [ "code-generation", "file-operations", "command-execution", "progress-monitoring" ], "dependencies": { "codebuddy-cli": ">=2.68.0" } } ``` --- ## 📚 API 文档 ### `execute(options)` 执行编程任务。 **参数:** ```typescript interface ExecuteOptions { task: string; // 任务描述 context?: { // 任务上下文 projectPath?: string; // 项目路径 techStack?: string; // 技术栈 files?: string[]; // 相关文件 }; options?: { // 执行选项 outputFormat?: 'json' | 'text'; // 输出格式 permissionMode?: 'default' | 'bypassPermissions'; // 权限模式 timeout?: number; // 超时时间(秒) }; } ``` **返回:** ```typescript interface ExecuteResult { status: 'success' | 'failed' | 'timeout'; filesModified: string[]; // 修改的文件列表 toolCalls: ToolCall[]; // 工具调用记录 reasoning: string[]; // 推理过程 duration: number; // 执行时长(秒) error?: string; // 错误信息 } ``` ### `onProgress(callback)` 订阅进度更新事件。 **参数:** ```typescript type ProgressCallback = (progress: { percentage: number; // 完成百分比 currentTask: string; // 当前任务描述 elapsedTime: number; // 已用时间(秒) estimatedTime?: number; // 预计剩余时间(秒) filesModified: string[]; // 已修改文件 toolCalls: number; // 已调用工具次数 }) => void; ``` ### `getStatus()` 获取当前任务状态。 **返回:** ```typescript interface TaskStatus { state: 'idle' | 'running' | 'completed' | 'failed'; taskId?: string; startTime?: Date; progress?: Progress; } ``` --- ## 🎯 使用示例 ### 示例1:创建新组件 ```javascript const codebuddy = require('./skill'); // 创建登录组件 const result = await codebuddy.execute({ task: '创建一个用户登录组件,包含用户名、密码输入框和登录按钮', context: { projectPath: '/path/to/vue-project', techStack: 'Vue 3 Composition API + TypeScript' } }); if (result.status === 'success') { console.log('组件创建成功!'); console.log('创建的文件:', result.filesModified); } ``` ### 示例2:修复 Bug ```javascript const codebuddy = require('./skill'); // 修复登录验证 Bug const result = await codebuddy.execute({ task: '修复用户登录时的验证逻辑,密码应该至少8位且包含数字和字母', context: { projectPath: '/path/to/project', files: ['src/views/Login.vue', 'src/utils/validator.ts'] } }); console.log('修复完成:', result.filesModified); ``` ### 示例3:监听长时间任务进度 ```javascript const codebuddy = require('./skill'); // 订阅进度 codebuddy.onProgress((progress) => { console.log(`[${progress.percentage}%] ${progress.currentTask}`); console.log(` 已修改 ${progress.filesModified.length} 个文件`); console.log(` 已执行 ${progress.toolCalls} 次操作`); console.log(` 用时 ${progress.elapsedTime}s`); }); // 执行长时间任务 const result = await codebuddy.execute({ task: '重构整个用户管理模块,使用更清晰的架构', context: { projectPath: '/path/to/project' }, options: { timeout: 600 // 10分钟超时 } }); ``` --- ## 🔍 进度监控原理 ### JSON 输出解析 CodeBuddy CLI 支持 `--output-format json` 输出结构化数据: ```bash codebuddy -p "任务描述" --output-format json --permission-mode bypassPermissions ``` **输出格式:** ```json { "status": "running", "tool_calls": [ { "tool": "write_to_file", "parameters": { "filePath": "src/Login.vue", "content": "..." }, "result": "success" } ], "files_modified": ["src/Login.vue"], "reasoning": [ "分析任务需求", "设计组件结构", "编写代码" ], "progress": { "percentage": 45, "current_task": "编写登录表单" } } ``` ### 进度解析流程 ```mermaid graph LR A[CodeBuddy CLI] -->|JSON Stream| B[Progress Monitor] B -->|Parse JSON| C[Progress Data] C -->|Emit Event| D[Event Callbacks] D -->|Update| E[Agent UI] ``` --- ## ⚙️ 高级配置 ### 自定义输出解析器 ```javascript const codebuddy = require('./skill'); // 自定义解析器 codebuddy.setOutputParser((jsonLine) => { // 自定义解析逻辑 return { percentage: jsonLine.progress?.percentage || 0, task: jsonLine.progress?.current_task || '处理中' }; }); ``` ### 超时和重试 ```javascript const result = await codebuddy.execute({ task: '复杂重构任务', options: { timeout: 1200, // 20分钟超时 retryCount: 3, // 失败重试3次 retryDelay: 5000 // 重试间隔5秒 } }); ``` --- ## 🐛 调试和日志 ### 启用详细日志 ```javascript const codebuddy = require('./skill'); // 启用调试模式 codebuddy.setDebugMode(true); // 所有 CLI 输出会被记录到控制台 const result = await codebuddy.execute({ task: '创建测试文件' }); ``` ### 查看执行日志 ```javascript // 获取最近的执行日志 const logs = codebuddy.getExecutionLogs(); console.log(logs); // [ // { time: '11:30:01', event: 'CLI_START', command: '...' }, // { time: '11:30:02', event: 'TOOL_CALL', tool: 'write_to_file' }, // { time: '11:30:05', event: 'CLI_END', status: 'success' } // ] ``` --- ## 🚨 错误处理 ### 错误类型 ```typescript enum CodeBuddyErrorType { CLI_NOT_FOUND = 'CLI_NOT_FOUND', // CodeBuddy CLI 未安装 INVALID_TASK = 'INVALID_TASK', // 无效的任务描述 TIMEOUT = 'TIMEOUT', // 执行超时 PERMISSION_DENIED = 'PERMISSION_DENIED', // 权限被拒绝 CLI_ERROR = 'CLI_ERROR' // CLI 执行错误 } ``` ### 错误处理示例 ```javascript try { const result = await codebuddy.execute({ task: '创建文件' }); } catch (error) { if (error.type === 'CLI_NOT_FOUND') { console.error('请先安装 CodeBuddy CLI'); } else if (error.type === 'TIMEOUT') { console.error('任务超时,请增加超时时间'); } else { console.error('执行失败:', error.message); } } ``` --- ## 📦 集成到 Agent ### Developer Agent 集成 ```javascript // developer/agent.js const codebuddy = require('codebuddy-coding'); class DeveloperAgent { async implementFeature(task) { // 使用 CodeBuddy 实现功能 const result = await codebuddy.execute({ task: task.description, context: { projectPath: this.projectPath, files: task.relatedFiles } }); return result; } } ``` ### Architect Agent 集成 ```javascript // architect/agent.js const codebuddy = require('codebuddy-coding'); class ArchitectAgent { async generateProjectScaffold(requirements) { // 使用 CodeBuddy 生成脚手架 const result = await codebuddy.execute({ task: `创建项目脚手架:${requirements}`, options: { permissionMode: 'bypassPermissions' } }); return result; } } ``` --- ## 🧪 测试 ### 运行测试 ```bash # 运行所有测试 npm test # 运行特定测试 npm test -- --grep "CLI Wrapper" ``` ### 测试覆盖 - ✅ CLI Wrapper 单元测试 - ✅ Progress Monitor 单元测试 - ✅ Integration 集成测试 - ✅ E2E 端到端测试 --- ## 📄 许可证 MIT License --- ## 🤝 贡献 欢迎提交 Issue 和 Pull Request! --- ## 📞 支持 如有问题,请联系: - GitHub Issues: [OpenClaw Repository] - Email: support@openclaw.ai --- **让每个 Agent 都拥有 AI 编程能力!** 🚀

Archived SourceRecently Updated
Coding

anti-sycophancy

Three-layer sycophancy defense based on ArXiv 2602.23971. Use /anti-sycophancy install to deploy all layers, or manage individually via install-claude-code / install-openclaw / uninstall / status / verify. Layer 1: CC-only hook; Layer 2: SKILL (cross-platform); Layer 3: CLAUDE.md (CC) / SOUL.md (OC).

Archived SourceRecently Updated
Coding

validation-rule-management

管理校验规则、规则组和校验场景的全流程操作。支持通过统一 CLI 工具快速执行 API 调用,自动处理参数解析、配置加载和错误提示。使用当用户需要进行校验规则管理、规则组维护、校验场景配置、启停操作或相关查询时,即使用户只说"帮我创建一条规则"或"查一下场景列表"也应触发。

Archived SourceRecently Updated