skill-creator
Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.
gemini-agent
>
generate-goal
Synthesize the current conversation into a single, self-contained `/goal` prompt that Claude Code can run to continue the agreed-upon work autonomously — so the user never has to hand-summarize a long discussion into a goal directive. Use when the user and Claude have converged on what to do next and the user wants to hand it off to the `/goal` command. Triggers include "generate a goal", "make/write a goal prompt", "turn this into a /goal", "hand this off to /goal", "goalify this", or "/generate-goal". Outputs the prompt to chat only; never writes files to disk.
git-worktree-prune
Safely prune stale git worktrees and local branches left over from past sessions, deleting only those with no unique work so no commit is ever lost. Use when a repo has accumulated worktrees and branches over many sessions and needs cleanup. Triggers include "prune worktrees", "delete stale branches", "clean up git worktrees/branches", "remove old worktrees", "which branches are safe to delete".
parallel-phases
End-to-end orchestrator for multi-task parallel work. Phase 1 (setup) gathers code intelligence, analyzes merge-conflict risks, groups tasks into maximally-parallel phases, writes enriched self-contained agent prompts, and generates PLAN.md + STATE.md. Phase 2 (execution) dispatches Opus agents in worktrees with a per-task review-gauntlet sub-skill pass and auto-fix commits, then a doc-update agent at completion. Asks explicit permission before starting autonomous execution. Resilient to /clear via file-backed state. Triggers on /parallel-phases, "parallelize these tasks", "run these in parallel phases".
root-cause-tracing
Use when errors occur deep in execution and you need to trace back to find the original trigger - systematically traces bugs backward through call stack, adding instrumentation when needed, to identify source of invalid data or incorrect behavior