@mmerterden/multi-agent-pipeline 20.6.0 → 20.7.0

package/pipeline/skills/shared/external/skill-creator/template.md

@@ -1,67 +1,79 @@
-# Skill skeletons
+# Skill templates
 
-Two skeletons - pick by layer. Both stay lean; push detail into one-level-deep bundled files.
+Two skeletons. Copy the one that matches the layer, fill it, then delete every line that fails the cut test. The toolkits use these frontmatter keys: `name`, `description`, `allowed-tools`, `user-invocable`, plus an optional `argument-hint`.
 
-Frontmatter fields used in this toolkit: `name`, `description` (required-quality), `user-invocable`, `argument-hint` (optional), `allowed-tools`.
-
----
-
-## Reference skill (descriptive - knowledge)
-
-No numbered procedures. Principles + pointers + heuristics.
+## Reference skill
 
 ```markdown
 ---
-name: <aspect-noun>          # e.g. navigation - NO platform prefix
-description: <what it covers + when to reach for it + trigger words. Say "iOS/SwiftUI" so the router can disambiguate.>
+name: {aspect-noun}
+description: "{What it covers} for {platform}. {When to reach for it}. Triggers: '{phrase}', '{phrase}', '{phrase}'."
 user-invocable: true
-allowed-tools: Read, Grep, Glob, Bash, Edit, Write
+allowed-tools: Bash, Read, Edit, Write, Glob, Grep
 ---
 
-# <name> - <short title>
+# {name} - {short title}
 
-> **Layer:** reference (knowledge)
-> **Use when:** <trigger>   **Hand off to:** <sibling skill>
+> **Layer:** reference (knowledge). **Use when:** {situation}. **Hand off to:** {sibling skill} for {adjacent concern}.
 
 ## Scope
-In / out, 1-2 lines.
+In: {one line}. Out: {one line}.
 
-## Invariants            <!-- non-negotiables + the why -->
-## Conventions & golden paths   <!-- point to canonical files; don't paste code -->
-## Decision rules        <!-- if X do A; when unsure, ask/verify -->
-## Pitfalls              <!-- known traps -->
-## Escape hatches        <!-- when this doesn't apply → which skill -->
-```
+## Invariants
+- **{Rule}.** {Why it holds; what breaks otherwise.}
 
----
+## Conventions and golden paths
+- {Convention} - see `{path/to/canonical/file}`. Point at the file; do not paste its code.
+
+## Decision rules
+- {Condition} -> {choice}.
 
-## Workflow skill (imperative - action)
+## Pitfalls
+- {Mistake} - {symptom} - {correct move}.
 
-Has a numbered procedure, but only for the deterministic/irreversible spine. Cite reference skills for judgment. Always include a verification gate.
+## Escape hatches
+- {Case this skill does not cover} -> {skill to use}.
+```
+
+## Workflow skill
 
 ```markdown
 ---
-name: <verb-phrase>         # e.g. branch-and-pr
-description: <what it does + when to run it + trigger words (iOS/SwiftUI).>
+name: {verb-phrase}
+description: "{What it does} for {platform}, with {guardrails}. Use when {situation}. Triggers: '{phrase}', '{phrase}'."
 user-invocable: true
-argument-hint: [<step> | ...]
+argument-hint: "{args}"
 allowed-tools: Bash, Read
 ---
 
-# <name> - <short title>
+# {name} - {short title}
 
-> **Layer:** workflow (action)
-> **Use when:** <trigger>   **Don't use when:** <hand off>
+> **Layer:** workflow (action). **Use when:** {situation}. **Don't use when:** {situation} -> {sibling skill}.
 
 ## Scope
-In / out, 1-2 lines.
-
-## Invariants            <!-- hard rules / guardrails -->
-## Conventions & golden paths   <!-- base branch, auth, canonical commands -->
-## Decision rules        <!-- branch when..., escalate when... -->
-## Procedures            <!-- IMPERATIVE: only deterministic/irreversible steps -->
-1. ...
-## Verification          <!-- IMPERATIVE: prove it worked (build/run/test) -->
+In: {one line}. Out: {one line}.
+
+## Invariants
+- **{Rule}.** {Why.}
+
+## Conventions and golden paths
+- Base branch: {branch}. Auth: {how}. Canonical commands: `{command}`.
+
+## Decision rules
+- {Condition} -> {choice}. Judgement calls cite `{reference skill}`.
+
+## Procedures
+1. {Deterministic, imperative step.}
+2. {Next step.}
+
+## Verification
+- Run `{build / test / check}` and confirm {observable result}. The skill is not done until this passes.
+
 ## Pitfalls
+- {Mistake} - {correct move}.
+
 ## Escape hatches
+- {Case outside scope} -> {skill to use}.
 ```
+
+A tool skill follows the workflow skeleton cut down to: when to use, the run command, a sub-command overview, and a link to its bundled `reference.md`.

package/pipeline/skills/shared/external/tailwind-css/SKILL.md

@@ -2,6 +2,7 @@
 name: tailwind-css
 description: "Tailwind CSS v4: utility-first styling, responsive design, dark mode, custom themes, component patterns, and cn() utility. Use when styling or reviewing UI with Tailwind: utilities, responsive design, dark mode, theming."
 metadata:
+  source: multi-agent-pipeline
   tags: [tailwind, css, styling, responsive, dark-mode, frontend]
   version: "2025.1"
 ---

package/pipeline/skills/shared/external/testing-backend/SKILL.md

@@ -2,6 +2,7 @@
 name: testing-backend
 description: "Backend testing: pytest patterns (Python), Jest patterns (Node.js), integration testing, database fixtures/factories, API contract testing. Use when writing or reviewing backend tests: unit, integration, database and API level."
 metadata:
+  source: multi-agent-pipeline
   tags: [testing, pytest, jest, integration, database, api-testing, backend]
   version: "2025.1"
 ---

package/pipeline/skills/shared/external/typescript-patterns/SKILL.md

@@ -2,6 +2,7 @@
 name: typescript-patterns
 description: "Modern TypeScript patterns: strict mode, utility types, generics, conditional types, type guards, discriminated unions, Zod validation. Use when writing or reviewing TypeScript types: generics, utility and conditional types, guards, strict mode."
 metadata:
+  source: multi-agent-pipeline
   tags: [typescript, types, generics, zod, validation, frontend, backend]
   version: "2025.1"
 ---

package/pipeline/skills/shared/external/vue-composition/SKILL.md

@@ -2,6 +2,7 @@
 name: vue-composition
 description: "Vue 3 Composition API: ref, reactive, computed, watch, composables, Pinia state management, Vue Router 4, script setup. Use when building or reviewing Vue 3 work with the Composition API, composables or Pinia."
 metadata:
+  source: multi-agent-pipeline
   tags: [vue, composition-api, pinia, vue-router, frontend, javascript]
   version: "2025.1"
 ---

package/pipeline/skills/shared/external/web-accessibility/SKILL.md

@@ -2,6 +2,7 @@
 name: web-accessibility
 description: "WCAG 2.2 AA web accessibility: ARIA roles/states, keyboard navigation, screen readers, color contrast, focus management, axe-core testing. Use when auditing or fixing a web UI against WCAG: ARIA, keyboard access, screen readers, contrast."
 metadata:
+  source: multi-agent-pipeline
   tags: [accessibility, a11y, wcag, aria, screen-reader, frontend]
   version: "2025.1"
 ---

package/pipeline/skills/shared/external/web-performance/SKILL.md

@@ -2,6 +2,7 @@
 name: web-performance
 description: "Web performance: Core Web Vitals (LCP, INP, CLS), image optimization, code splitting, caching strategies, bundle analysis, Lighthouse. Use when a page is slow and Core Web Vitals, images, bundles or caching have to be addressed."
 metadata:
+  source: multi-agent-pipeline
   tags: [performance, core-web-vitals, lighthouse, optimization, frontend, caching]
   version: "2025.1"
 ---

package/pipeline/skills/shared/external/web-testing/SKILL.md

@@ -2,6 +2,7 @@
 name: web-testing
 description: "Web testing: Vitest unit tests, React Testing Library, Playwright E2E, MSW API mocking, testing hooks and async operations. Use when writing or reviewing web tests with Vitest, Testing Library, Playwright or MSW."
 metadata:
+  source: multi-agent-pipeline
   tags: [testing, vitest, playwright, react-testing-library, msw, e2e, frontend]
   version: "2025.1"
 ---