Update to EngineeringSuite

{
  "name": "engineering-suite",
  "description": "Use whenever the EngineeringSuite plugin is selected, the user writes EngineeringSuite, @EngineeringSuite, or a command like EngineeringSuite /research. Routes explicit slash commands to the exact installed upstream skill before any response or action, and otherwise selects the smallest relevant set of installed skills.",
  "included_files": [],
  "skill_md_contents": "---\nname: engineering-suite\ndescription: Use whenever the EngineeringSuite plugin is selected, the user writes EngineeringSuite, @EngineeringSuite, or a command like EngineeringSuite /research. Routes explicit slash commands to the exact installed upstream skill before any response or action, and otherwise selects the smallest relevant set of installed skills.\nlicense: MIT\n---\n\n# EngineeringSuite Router\n\nEngineeringSuite contains the complete vendored skill sets from AI Hero / Matt Pocock, Ponytail, Addy Osmani Agent Skills, and Taste Skill. This router does not replace those skills. It selects and invokes them.\n\n## Mandatory routing rule\n\nBefore answering or taking any action, inspect the user's message for an Engineering Suite command.\n\nAccepted forms include `EngineeringSuite /research ...`, `@EngineeringSuite /research ...`, and a slash command supplied while this plugin is explicitly selected.\n\nWhen an explicit command is present:\n\n1. Resolve the command before doing anything else.\n2. Invoke the exact installed skill named by the route.\n3. Read and follow that target skill as written. Do not summarize it into a substitute workflow.\n4. The explicit route overrides automatic routing. Do not silently replace it with a different methodology.\n5. If several commands are supplied, invoke them left-to-right unless their own instructions establish a different process order.\n\n## Exact-name routing\n\nEvery installed upstream skill is addressable directly by its `name:` value:\n\n`EngineeringSuite /<skill-name> <request>` → invoke the exact installed skill `<skill-name>`.\n\nExamples:\n\n- `EngineeringSuite /research Compare two current libraries` → `research`\n- `EngineeringSuite /tdd Implement this behavior` → `tdd`\n- `EngineeringSuite /code-review Review this diff` → `code-review`\n- `EngineeringSuite /design-taste-frontend Redesign this page` → `design-taste-frontend`\n- `EngineeringSuite /ponytail-audit Audit this codebase` → `ponytail-audit`\n- `EngineeringSuite /security-and-hardening Threat-model this endpoint` → `security-and-hardening`\n\n## Convenience aliases\n\nThese shorter routes are aliases only; the target remains the original installed skill:\n\n- `/research` → `research`\n- `/review` → `code-review`\n- `/debug` → `diagnosing-bugs`\n- `/tdd` → `tdd`\n- `/spec` → `to-spec`\n- `/tickets` → `to-tickets`\n- `/implement` → `implement`\n- `/prototype` → `prototype`\n- `/wayfinder` → `wayfinder`\n- `/grill` → `grill-with-docs`\n- `/handoff` → `handoff`\n- `/architecture` → `improve-codebase-architecture`\n- `/taste` → `design-taste-frontend`\n- `/redesign` → `redesign-existing-projects`\n- `/minimal` → `minimalist-ui`\n- `/brutalist` → `industrial-brutalist-ui`\n- `/ponytail` → `ponytail`\n- `/audit` → `ponytail-audit`\n- `/debt` → `ponytail-debt`\n- `/gain` → `ponytail-gain`\n- `/quality` → `code-review-and-quality`\n- `/simplify` → `code-simplification`\n- `/security` → `security-and-hardening`\n- `/performance` → `performance-optimization`\n- `/ship` → `shipping-and-launch`\n- `/plan` → `planning-and-task-breakdown`\n- `/ui-engineering` → `frontend-ui-engineering`\n\n## Source-qualified routing\n\nSource prefixes are optional and are useful for clarity:\n\n- `EngineeringSuite /aihero:research` → `research`\n- `EngineeringSuite /aihero:tdd` → `tdd`\n- `EngineeringSuite /ponytail:audit` → `ponytail-audit`\n- `EngineeringSuite /addy:test-driven-development` → `test-driven-development`\n- `EngineeringSuite /taste:frontend` → `design-taste-frontend`\n- `EngineeringSuite /taste:redesign` → `redesign-existing-projects`\n\nIf a source-qualified command does not map to a skill from that source, report the mismatch and suggest exact valid names. Do not guess.\n\n## Automatic routing when there is no slash command\n\nChoose the smallest useful set. Prefer a process skill first, then a domain skill only when needed.\n\n- Open-ended engineering question or current technical investigation: `research`.\n- Unclear feature or requirements: `grill-with-docs`, then `to-spec` if the user wants a durable spec.\n- Implementation from an accepted spec: `implement`; use `tdd` when behavior changes.\n- Bug investigation: `diagnosing-bugs`, then `tdd` for the fix.\n- Code review: `code-review`; add `code-review-and-quality` when production-quality gates are explicitly relevant.\n- Architecture exploration: `wayfinder` or `improve-codebase-architecture` depending on whether the goal is discovery or refactoring.\n- Simplification / YAGNI / unnecessary complexity: `ponytail`, `ponytail-review`, or `code-simplification` according to the request.\n- Frontend creation: `design-taste-frontend`; existing UI redesign: `redesign-existing-projects`; implementation quality concerns: `frontend-ui-engineering`.\n- Security, performance, observability, CI/CD, release, API design, migrations: use the matching Addy skill only when the task actually needs it.\n\nDo not invoke all skill families by default. Explicit user commands always win.\n\n## Help\n\nFor `EngineeringSuite /help` or `/skills`, read `../../references/command-catalog.md` and present the relevant routes grouped by source.\n"
}