Browse AI and ML workflow skills for model integration, prompt engineering, evaluations, and LLM automation across major IDEs.
This directory brings installable AI Agent skills into one place so you can filter by search, category, topic, and official source, then install them directly into Claude Code, Cursor, Windsurf, and other supported environments.
Debug AI traces, find exceptions, analyze sessions, and manage prompts via Langfuse MCP. Use when debugging AI pipelines, investigating errors, analyzing latency, managing prompt versions, or setting up Langfuse. Triggers on langfuse, traces, debug AI, find exceptions, what went wrong, why is it slow, datasets, evaluation sets.
Creates production-ready pull requests with automatically updated AGENTS.md and project docs. Use when the user asks to create a PR, make a PR, is ready to merge, has finished a feature and wants a PR, or says theyre done for the day with changes to commit.
Run ML training, LLM inference, and ComfyUI workflows on remote NVIDIA GPUs (A100, H100, RTX 4090). Cloud GPU compute with smart file sync — prefix any command with gpu to run it remotely.
Run Agent Workbench (awb) interactive UI components from Codex or the CLI. Use when the user needs a confirmation, checklist, input, table, code or markdown view, or a visual playground. Always follow interactive runs with awb wait and the returned ID. Use tmux for long-running commands (see awb --help).
Manages project progress tracking and maintains a chronological log of completed tasks, decisions, and updates. Use when completing milestones, making architectural decisions, or documenting project evolution. Creates and updates LOG.md file based on CLAUDE.md context.
The api-rules skill provides a Python programming assistant for LLM evaluation tasks, utilizing OpenAI API and libraries like Pandas and NumPy. It benefits developers working with LLMs.
Shared utilities for Gemini-based skills including API key management, Vertex AI configuration, and client helpers. Use as dependency for skills requiring Gemini API access.
Shared utilities for Gemini-based skills including API key management, Vertex AI configuration, and client helpers. Use as dependency for skills requiring Gemini API access.
This skill should be used when the user asks to build a complete app, develop a software project from scratch, create a website with full workflow, manage a development team, automate software development from requirements to deployment, or coordinate multiple agents for software development. Integrates Superpowers frameworks: brainstorming for requirement exploration, git-worktrees for workspace isolation, subagent-driven-development for task execution, parallel-agent-dispatch for testing, TDD for code quality, and two-stage code review for specification compliance. Also includes integration testing with Chrome DevTools MCP for end-to-end validation. 🆕 v6.6.0: Complete workflow automation with auto/interactive modes, Git worktree for seamless project creation, and Callstack React Native best practices integration.
Rewrite and clarify a users goal before starting a multi-agent workflow; produce a tighter goal plus constraints, acceptance criteria, assumptions, and any clarifying questions.
Review and rate code implementations against issue cards and ExecPlans using git diffs, AGENTS.md, and PLANS.md. Use when a user asks for an implementation review, compliance check, regression risk assessment, or to validate changes against one or more issue cards or ExecPlans.
CLAUDE.md/AGENTS.md等のコアルールファイルの保守・更新を対話で行い、3環境同期と整合性を維持したまま反映する。CLAUDE.md更新、コアルール変更、ルール追加、パス辞書更新、ワークフロー索引更新を依頼されたときに使用する。