agent-audit
Audit agents spawned in the current/last run against the agent-selection taxonomy
Chain the full Superpowers process flow — brainstorm → plan → worktree → build → review → finish — on one slash command
> /plugin marketplace add heymegabyte/claude-skillsHow it fires
How this command gets triggered: by you, by Claude, or both.
/processContext preview
What this command does when you run it.
Chain the full Superpowers process flow — brainstorm → plan → worktree → build → review → finish — on one slash command
description: Chain the full Superpowers process flow — brainstorm → plan → worktree → build → review → finish — on one slash command argument-hint: <task-description> disable-model-invocation: true
<!-- <SUBAGENT-STOP>: skip this skill when running inside a subagent. Meta-skills must not leak into spawned subagent contexts. --> <SUBAGENT-STOP/>
Run the full process chain from `20-superpowers` (vendored from obra/Superpowers, MIT), compressed to house style. This is the keystone command — from a one-line task description to a merged, deployed, verified change.
1. `20/brainstorming` — explore intent, requirements, and design BEFORE any code. Present design → get approval → write `docs/superpowers/specs/YYYY-MM-DD-<topic>-design.md`. 2. `20/writing-plans` — turn the approved design into a written plan with bite-sized verifiable steps. 3. `20/using-git-worktrees` — isolate the workspace (`EnterWorktree` or `git worktree`). Skip if already in one. 4. `20/subagent-driven-development` — one implementer subagent per independent task, task review after each, whole-branch review at the end. 5. `20/requesting-code-review` — dispatch the `code-reviewer` agent + Agent Diversity Review gate (`[[agent-selection]]`). 6. `20/receiving-code-review` — verify feedback technically, then act. Push back if wrong. 7. `20/finishing-a-development-branch` — merge + push + clean up. Auto-push per `[[no-staging-doctrine]]`.
14-category autonomous product-building OS for 32+ AI coding tools. One-line prompts → deployed products.
Repo: heymegabyte/claude-skills
Audit agents spawned in the current/last run against the agent-selection taxonomy
Run the Agent Diversity Review gate and emit the result table
Meta-analyze the effectiveness of a /loop arc — per-iteration metrics, LOC delta trend, saturation detection, and a keep/lengthen/delete recommendation.
Audit the rules/ directory for missing foundational principles; output gap list with priority and justification
Validate ~/.claude/settings.json hooks block — event names, file existence, executability, matcher syntax; --fix repairs common issues
Catch Resend-class bug (isError: false on HTTP 4xx/5xx) across all MCP server tool handlers