Find the perfect capability for your agent.
Create isolated git worktrees for ALL coding tasks with issue linking (discovery: implementation-clarity). Evaluate at implementation-clarity when task modifies code files. YOU MUST use for feature development, bug fixes, refactoring - no exceptions. Coding without worktree = no audit trail, no issue linking, no auto-close.
Transforms repositories into structured plain-text digests optimized for LLM consumption. Use when analyzing GitHub repositories, digesting codebases, or ingesting git repos for AI analysis.
Safely rebase the current feature branch on top of the latest origin/main. Use when preparing a branch for PR, UAT, or release.
Navigate a local emacs-twist/twist.nix checkout and point to relevant files for questions about twist.nix configuration, flake outputs, library functions, Home Manager integration, package build flow, or tests. Use when a user asks where something is defined or how twist.nix is wired and the answer should reference code rather than reproduce it. Ask for the repo path if it is not available.
Generate semantic commit messages and create git commits. Handles OneDrive mmap errors automatically. Use when user wants to commit, needs a commit message, or git commit fails.
This skill provides comprehensive guidance for learning and using GitHub, from fundamental concepts through practical workflows. Designed for complete beginners, this skill should be used when teaching GitHub concepts, explaining git workflows, debugging git/GitHub issues, or guiding through practical GitHub tasks like creating branches, commits, pull requests, and merging code.
Use when feature branch is complete and reviewed, ready to merge to main - squashes multiple commits into single clean commit in main using git merge --squash
Enhances Claude Code branch creation with mainline detection, conventional naming, and base branch awareness. Integrates with standard git workflows—adds automatic base branch detection, naming convention inference, and remote sync before creating branches. Use for 'create branch', 'new branch', 'start branch for', 'feature branch', 'work on'.
Safely sync a branch with origin/main using fetch + rebase, with an optional "accept current" conflict policy.
Provides comprehensive project folder structure design guidelines and best practices. Defines standard directory organizations for diverse project types including monorepos, web frameworks, backend services, libraries, and extensions. Ensures scalable, maintainable architecture through consistent file organization patterns. Specializes in separation of concerns, modular architecture, and tooling integration. Use when: designing new project structures, organizing monorepo workspaces with tools like Turborepo/Nx, structuring NestJS backend projects, organizing React/Next.js frontend applications, designing Go service architectures, creating NPM package structures, organizing VSCode extension projects, structuring Chrome extension codebases, planning directory hierarchies, migrating legacy project structures, or establishing code organization conventions for teams.
This skill should be used when the user needs to review and respond to GitHub Pull Request comments. It automatically fetches review comments (both inline code comments and PR-level comments) from GitHub Codex, Claude Code, or human reviewers, organizes them by priority, and lets the user select which items to address using AskUserQuestion. After the user selects items, the main agent automatically implements the fixes.
Strict detection and removal of unrelated files before PR creation
Maintain README.md files with GitHub advanced formatting (tables, badges, diagrams, collapsible sections)
Create atomic commits grouped by logical features with README.md and CLAUDE.md awareness.
Guide for creating and maintaining repository-local CLAUDE.md files that capture project-specific context, patterns, and conventions. Use when initializing repositories for Claude Code, updating project memory, or understanding effective repository memory structure.
Systematic codebase exploration using tree-sitter AST analysis and repomix semantic search. Use when: (1) User asks to explore/analyze/find/search codebase, (2) Looking for functions/classes/symbols/patterns, (3) Understanding project structure, (4) Tracing dependencies/usage, (5) ANY code analysis task. CRITICAL: NEVER use basic filesystem tools (find, grep, Glob, Read multiple files) when tree-sitter MCP or repomix tools are available.
Best practices for project documentation including README, CONTRIBUTING, CHANGELOG, and Markdown files. Use when updating documentation, writing README sections, or maintaining project docs.
