Co-Pilot / 辅助式
更新于 a month ago

ai-toolbox

Ddesplega-ai
0.0k
desplega-ai/ai-toolbox
76
Agent 评分

💡 摘要

AI工具箱提供了一套用于AI和开发者生产力的工具,包括各种任务的插件。

🎯 适合人群

寻找生产力工具的AI开发者需要代码贡献跟踪的软件工程师管理知识和笔记的研究人员协调多代理任务的项目经理管理发票和任务的自由职业者

🤖 AI 吐槽:看起来很能打,但别让配置把人劝退。

安全分析中风险

风险:Medium。建议检查:是否执行 shell/命令行指令;是否发起外网请求(SSRF/数据外发);文件读写范围与路径穿越风险;依赖锁定与供应链风险。以最小权限运行,并在生产环境启用前审计代码与依赖。

AI Toolbox

Monorepo of AI and developer tools by desplega.ai. Each subdirectory is a standalone tool with its own README.

Repository Structure

| Tool | Description | Stack | |------|-------------|-------| | ai-tracker | Track AI vs human code contributions | Python, uv, SQLite | | brain | Personal knowledge management with semantic search | TypeScript, Bun, SQLite | | cc-hooks | macOS notifications for Claude Code | Bash | | cc-notch | Menu bar cost tracker (SwiftBar) | Shell | | cc-plugin | Claude Code plugins | YAML/Markdown | | cc-what | SDK & CLI for analyzing Claude Code usage data | TypeScript, Bun | | dns | DNS TXT record query utility | - | | file-review | File review tool | - | | hive | macOS app for Claude Code sessions | TypeScript, Electron, Vite | | hn-sql | HN data with Parquet + SQL | - | | invoice-cli | Invoice email fetcher | - | | thoughts | Research notes & plans (via /desplega:*) | Markdown | | willitfront.page | HN analysis with natural language | - | | wts | Git worktree manager | TypeScript, Bun |

Thoughts

The thoughts directory contains research notes and implementation plans generated via Claude Code skills from the base plugin:

  • /desplega:research - Document codebase state with research notes
  • /desplega:create-plan - Create detailed implementation plans
  • /desplega:implement-plan - Execute plans from a plan file

Claude Plugins

Installation

From inside Claude Code:

/plugin marketplace add desplega-ai/ai-toolbox

Or from the terminal:

claude plugin marketplace add desplega-ai/ai-toolbox

Available Plugins

| Plugin | Description | |--------|-------------| | desplega | Base agentic coding patterns (research, planning, implementation) | | agent-swarm | Multi-agent coordination | | wts | Git worktree manager | | brain | Personal knowledge management integration | | file-review | File review with inline comments | | remarkable | reMarkable tablet integration (get, put, ls) |

/plugin install desplega@desplega-ai-toolbox /plugin install agent-swarm@desplega-ai-toolbox /plugin install wts@desplega-ai-toolbox /plugin install brain@desplega-ai-toolbox /plugin install file-review@desplega-ai-toolbox /plugin install remarkable@desplega-ai-toolbox

Bash Install

You can also install from the terminal:

# Scope can be 'user' (default), 'project' or 'local' claude plugin install desplega@desplega-ai-toolbox --scope user claude plugin install agent-swarm@desplega-ai-toolbox --scope user claude plugin install wts@desplega-ai-toolbox --scope user claude plugin install brain@desplega-ai-toolbox --scope user claude plugin install file-review@desplega-ai-toolbox --scope user claude plugin install remarkable@desplega-ai-toolbox --scope user

Plugin structure: cc-plugin/{base,swarm,wts,brain,file-review,remarkable}/ with hooks, skills, and agents.

Key Tools

ai-tracker (Python):

uvx cc-ai-tracker install # Install hooks uvx cc-ai-tracker stats # View AI/human contribution stats

brain (TypeScript/Bun):

npm install -g @desplega.ai/brain brain init && brain add "My first note" brain search "ideas" # Semantic search

cc-what (TypeScript/Bun):

npm install -g @desplega.ai/cc-what cc-what # Summary of Claude Code usage cc-what stats --daily # Daily activity stats cc-what costs --models # Cost breakdown by model

wts (TypeScript/Bun):

npm install -g @desplega.ai/wts wts init && wts create feature --new-branch

Development

Each tool is independent - check its README for specific setup. General pattern:

  • Python tools: uv or uvx
  • TypeScript tools: bun or pnpm
  • Plugins: YAML/Markdown configs in cc-plugin/

New tools: Create folder with minimal slug, add README.md.

五维分析
清晰度8/10
创新性6/10
实用性9/10
完整性8/10
可维护性7/10
优缺点分析

优点

  • 多样化的工具集用于各种任务
  • 支持AI和开发者需求
  • 通过命令行轻松安装

缺点

  • 文档可以更详细
  • 工具是独立的,需要单独设置
  • 某些插件的信息有限

相关技能

pytorch

S
toolCode Lib / 代码库
92/ 100

“它是深度学习的瑞士军刀,但祝你好运能从47种安装方法里找到那个不会搞崩你系统的那一个。”

agno

S
toolCode Lib / 代码库
90/ 100

“它承诺成为智能体领域的Kubernetes,但得看开发者有没有耐心学习又一个编排层。”

nuxt-skills

S
toolCo-Pilot / 辅助式
90/ 100

“这本质上是一份组织良好的小抄,能把你的 AI 助手变成一只 Nuxt 框架的复读机。”

免责声明:本内容来源于 GitHub 开源项目,仅供展示和评分分析使用。

版权归原作者所有 desplega-ai.