om-superpowers

Open Mercato architecture skill — gap analysis, BC/risk analysis, skill-corpus audits, upstream bug triage, and orchestrating spec writing or implementation. Routes via Task Router. Triggers — "does OM do X", "analyze spec", "BC analysis", "review skills", "OM upstream bug", "write specs", "implement specs".

Diagnose and fix common issues in Open Mercato standalone apps — errors, modules not loading, widgets missing, failed migrations, build errors. Triggers — "error", "not working", "broken", "fix", "debug", "404", "500", "module not found". When the trace points into `node_modules/@open-mercato/*`, STOP and route to om-cto upstream-bug-triage before proposing any fix or PR.

OM design-system enforcement + backend admin page builder — flag hardcoded colors/typography/raw HTML controls, build DS-compliant CRUD/data-table/form pages, run DS health checks. Triggers — "DS review", "build admin page", "data table", "CRUD interface", "migrate colors", "DS health".

Implement OM specs end-to-end (or by phase) — coordinated subagents, code-review + tests gated per phase. Loads references for module scaffolding, data-model design, system extension (UMES), integration builders. Triggers — "implement spec", "create module", "design entity", "extend module", "build integration".

Use when defining business requirements for a new app or feature — BEFORE any spec or code exists. Triggers on "what should we build", "define requirements", "create app spec", or when user stories lack success criteria. Also use when reviewing existing specs for gaps.

Use when defining UI architecture, reviewing page layouts, navigation, widget placement, or user flow. Triggers — "what should the user see", "where does this go in the UI", "what pages do we need", or App Spec §3.5 UI Architecture reviews.

Resume an in-progress PR started by auto-create-pr. Claims the PR, checks out the branch in an isolated worktree, picks up from the first unchecked Progress step, and runs the same validation gate + label discipline. Usage — /auto-continue-pr <PR-number>.

Execute an autonomous agent task end-to-end and deliver it as a GitHub PR against `develop`. Drafts an execution plan with a Progress checklist, implements phase-by-phase in an isolated worktree with incremental commits, runs the validation gate (typecheck, unit tests, i18n, build). Resumable via auto-continue-pr.

Review or re-review a GitHub PR by number in an isolated worktree. Fetch the PR, run the code-review skill, submit approve/request-changes, manage labels. Offers an optional autofix flow that iterates conflicts, fixes, tests, typecheck, re-review until merge-ready. Usage — /auto-review-pr <PR-number>.

Review code changes for Open Mercato compliance — architecture, security, conventions, quality. Covers module structure, naming, data security, UI patterns, event/cache/queue rules, and anti-patterns. Use for PR reviews, audits, diffs, or commit reviews.

Run and create QA integration tests (Playwright TypeScript) — execute the full suite, convert optional markdown scenarios, generate tests from specs or features. Triggers — "run integration tests", "test this feature", "create test for", "QA tests".