๐งฐDeveloper Tools Skills
Browse the best Developer Tools skills for Claude Code, Codex, Cursor and other AI agents.
7,129 skills ยท Page 10 of 298
claude-cookbooks
โ
5Look up examples and patterns from the official Anthropic Claude Cookbooks repository. Use when you need reference implementations, code examples, or best practices for Claude integrations, tool use, agents, prompting techniques, or API usage.
directus-development-workflow
โ
5Complete development setup: scaffolding, TypeScript, testing, CI/CD, Docker, deployment, and best practices
mcp-builder
โ
2Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
llm-application-dev
โ
1Building applications with Large Language Models - prompt engineering,
git-conventional-commits
โ
1Apply when writing commit messages to maintain consistent, readable git history that enables automated changelog generation.
smith-dev
Development workflow standards and code quality requirements. Use when initializing projects, running quality checks, or managing agent tasks. Covers pre-commit checks, task decomposition, and script organization patterns.
claude-code-features
Explains Claude Code's extensibility features (skills, slash commands, hooks, MCP servers, settings, CLAUDE.md). Use when the user asks about Claude Code capabilities, how to set up custom commands, how to create skills, how to configure hooks, MCP server integration, or any meta-questions about Claude Code itself like "can you do X" or "how do I make you do Y".
searcher
Use this skill when the user wants to search for files by text content, find files containing specific text, or get a list of file paths related to a search term. The skill searches the codebase and returns matching file paths with context.
git-version-control
Git workflows, branching, collaboration, and code management
mac-automation
This skill should be used when the user requests Mac automation via AppleScript, including: Mail operations, Reminders/Calendar management, Safari control, Finder operations, system controls (volume, brightness, notifications, app launching), third-party apps (development tools, task management, media players), or mentions "AppleScript" or "automate Mac".
cua-cloud
Comprehensive guide for building Computer Use Agents with the CUA framework. This skill should be used when automating desktop applications, building vision-based agents, controlling virtual machines (Linux/Windows/macOS), or integrating computer-use models from Anthropic, OpenAI, or other providers. Covers Computer SDK (click, type, scroll, screenshot), Agent SDK (model configuration, composition), supported models, provider setup, and MCP integration.
anthropic-dev-tools-mcp-builder
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
git-workflow
Expert git workflow management. Use when working with git operations, branching, merging, resolving conflicts, or managing commits.
terminal
Execute shell commands in persistent, stateful VT100 terminal sessions. WHEN: User needs to "run a command", "execute shell", "build project", "run tests", "deploy", system administration. WHEN NOT: Simple file operations (use fs_* tools), reading known files (use fs_read_file).
epcp-workflow
Apply the Explore โ Plan โ Code โ Commit workflow for tasks like implementing features, fixing bugs, refactors, or adding integrations. Use this when the user wants changes in a repo and wants higher reliability (read first, plan then commit/PR). Emphasize subagents for investigation during Explore, and use "think / think hard / think harder / ultrathink" during Plan when alternatives exist.
ultrathink
่ค้ใชๅ้กใซๅฏพใใฆ่ชๅ็ใซๆทฑใๆ่ใขใผใใซๅใๆฟใใใ้ฃ่งฃใชใใฐใใขใผใญใใฏใใฃ่จญ่จใใใใฉใผใใณในๅ้กใชใฉใ้ๅธธใฎๆ่ใงใฏ่งฃๆฑบใ้ฃใใๅ้กใๆคๅบใใใ่ชๅ็ใซๆทฑๆ่ใขใผใใๆๅนๅใ
skill-developer
โ
20.1kCreate and manage Claude Code skills following Anthropic best practices. Use when creating new skills, modifying skill-rules.json, understanding trigger patterns, working with hooks, debugging skill activation, or implementing progressive disclosure. Covers skill structure, YAML frontmatter, trigger types (keywords, intent patterns, file paths, content patterns), enforcement levels (block, suggest, warn), hook mechanisms (UserPromptSubmit, PreToolUse), session tracking, and the 500-line rule.
artifacts-builder
โ
17.4kSuite of tools for creating elaborate, multi-component claude.ai HTML artifacts using modern frontend web technologies (React, Tailwind CSS, shadcn/ui). Use for complex artifacts requiring state management, routing, or shadcn/ui components - not for simple single-file HTML/JSX artifacts.
mcp-builder
โ
17.4kGuide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
huggingface-accelerate
โ
15.4kSimplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpeed/FSDP/Megatron/DDP. Automatic device placement, mixed precision (FP16/BF16/FP8). Interactive config, single launch command. HuggingFace ecosystem standard.
awq-quantization
โ
15.4kActivation-aware weight quantization for 4-bit LLM compression with 3x speedup and minimal accuracy loss. Use when deploying large models (7B-70B) on limited GPU memory, when you need faster inference than GPTQ with better accuracy preservation, or for instruction-tuned and multimodal models. MLSys 2024 Best Paper Award winner.
fine-tuning-with-trl
โ
15.4kFine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.
openrlhf-training
โ
15.4kHigh-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2ร faster than DeepSpeedChat with distributed architecture and GPU resource sharing.
implementing-llms-litgpt
โ
15.4kImplements and trains LLMs using Lightning AI's LitGPT with 20+ pretrained architectures (Llama, Gemma, Phi, Qwen, Mistral). Use when need clean model implementations, educational understanding of architectures, or production fine-tuning with LoRA/QLoRA. Single-file implementations, no abstraction layers.