audiogen

Generate game audio (music, voices, sound effects) via ElevenLabs.

Generate a project briefing: worktree status, open checkboxes, recent commits. Modes: summary (default), report, verify, current, worktrees. Period: 1h, 6h, 24h, 2d, 7d.

Safe commit workflow with optional scope hint. Inventories all changes, classifies related vs. unrelated files, traces dependencies, protects other agents' work, and optionally pushes or lands worktree commits. Usage: /commit [pr] [scope] [push|land]

Lightweight task dispatcher for ad-hoc work: documentation, examples, refactoring, content updates. Supports scheduling with every/now/next/stop. Usage: /do <description> [worktree] [push] [pr] [every SCHEDULE] [now] | stop | next.

Audit and fix documentation gaps: block library panel entries, example models, getting-started guides, block library, presentations, README updates. Also handles newsletter entries (/doc newsletter). Use when the user asks to "write a newsletter entry", "add to the newsletter", or "update the newsletter". Usage: /doc [blocks|examples|newsletter|<description>]

Draft a high-quality plan through iterative adversarial review. Multiple rounds of research, drafting, review, devil's advocate, and refinement until the plan converges. Output is a plan file ready for /run-plan. Usage: /draft-plan [output FILE] [rounds N] <description...>

Orchestrate a batch bug-fixing sprint. Supports scheduling with every/now/next/stop. Use sync to update trackers and verify/close already-fixed issues. Use plan to draft plans for skipped issues. Usage: /fix-issues N [focus] [auto] [every SCHEDULE] [now] | sync | plan [auto] | stop | next.

Review ALL unreported sprint results: walk through manual verifications, land fixes to main, close GitHub issues, update trackers, and clean up worktrees. Covers the current SPRINT_REPORT.md AND any landed-but-unclosed issues from previous sprints.

Deep debugging for complex bugs. Enforces a disciplined workflow: reproduce, trace, state root cause, fix, verify. The agent must PROVE it understands the root cause before writing any fix.

Block-diagram editor manual testing recipes for playwright-cli. Covers common UI operations (adding blocks, connecting ports, running simulations, editing parameters) using real mouse/keyboard events. Use when told to "test manually", "test in the browser", or "verify with playwright-cli".

Plan dashboard and batch executor. View plan status, find the next ready plan, or work through plans automatically. Usage: /plans [rebuild | next | work N [auto] [every SCHEDULE]]

QE audit: check recent commits for test coverage gaps, or bash/stress-test features to find bugs. Supports scheduling with every/now/next/stop. Usage: /qe-audit [bash [area]] [every SCHEDULE] [now] | stop | next.

Refine an in-progress plan by reviewing remaining phases against completed work. Dispatches adversarial reviewer and devil's advocate agents to find stale references, invalidated assumptions, and specification gaps — then refines remaining phases until convergence. Completed phases are NEVER modified. Appends a Drift Log and Plan Review section. Usage: /refine-plan <plan-file> [rounds N]

Full pipeline: decompose a broad goal into sub-plans, draft each with adversarial review, then execute all of them autonomously. One command, walk away. Usage: /research-and-go <description>

Decompose a broad goal into a sequence of executable sub-plans. Researches the domain, identifies sub-problems and dependencies, then produces a meta-plan where each phase delegates to /run-plan. Usage: /research-and-plan [output FILE] <description...>

Review exported feedback JSON from the in-app feedback panel, evaluate each pending entry, and selectively file GitHub issues via gh CLI. Use when the user says "review feedback", "triage feedback", or "file feedback issues".

Execute the next phase of a plan document: parse phases and status, dispatch implementation in a worktree, verify with a separate agent, update progress tracking, write reports/plan-{slug}.md, and optionally auto-land to main. Can self-schedule recurring runs via cron. Use `next` to check schedule, `stop` to cancel.

Install or update Z Skills supporting infrastructure (CLAUDE.md rules, hooks, scripts)

Verify all recent changes: review diffs, check that appropriate unit/e2e tests exist and correctly test the changes, run all tests, manually verify UI changes with playwright-cli, fix any problems found, re-verify until clean, then report results with recommendations.