Compile Central Alignment

Compile Central Alignment

Invoke as $compile-central-alignment.

Generate a browsable alignment/index.html linking to every alignment page in the current repository. Works in any repo with an alignment/ directory.

Process

1. Locate alignment/ at the project root; stop if absent.
2. Scan alignment/*.html (exclude index.html), extract per-file: <title>, first <h1>, first <p class="meta">, data-alignment-category from <html> element, modification date.
3. Group and sort:
   • Assign category per page: data-alignment-category attribute first; else match filename against prefix lists:
     • research: devtool-positioning, devtool-adoption, devtool-monetization, devtool-user-map, devtool-workflow, devtool-dx-journey, devtool-integration-map, deep-research, repo-glossary
     • product-design: idea-scope-brief, fork-idea-branch, animation-design-planner, skills-showcase, skills-inventory
     • utility: investigate, session-triage, prompt-history-backfill
     • qa-meta: expert-review, benchmark, targeted-skill-builder, devtool-docs-audit
     • ops-analysis: analyze-sessions, canonical-workflow, run-batch
     • Default: research
   • Group by category in fixed order: research → product-design → utility → qa-meta → ops-analysis
   • Within each category, sort by modification date descending
4. Generate alignment/index.html — self-contained HTML with:
   • Visual style matching existing alignment pages (same CSS variables, fonts, layout)
   • Category sections: each non-empty category gets an <h2> with display name and count (e.g., "Research (7)"), followed by its card grid
   • Display names: Research, Product Design & Spec, Utility & Maintenance, QA & Meta-Skill Improvement, Ops & Session Analysis
   • Card grid with linked titles, descriptions, dates
   • Text filter input searching across all categories, hiding empty category headings
   • Total page count and generation timestamp
   • All relative links, no external dependencies
   • Skip categories with zero pages
5. Verify alignment/index.html was written and still works as a direct file:// page.
6. Open or focus the generated index with:
   • npx skillpacks alignment pages open alignment/index.html --browser auto
   • In this source checkout only, if the packaged CLI is unavailable, fall back to node scripts/open-html-page.mjs alignment/index.html --browser auto.
7. Report the opener script status exactly as one of focused, opened, fallback-opened, blocked, or failed.
   • blocked means the browser-open attempt was unavailable or blocked; it does not fail this skill when alignment/index.html was generated and verified.

Constraints

• Do not modify existing alignment pages — only create or overwrite alignment/index.html.
• Do not commit or push alignment/index.html by default — it is a local convenience artifact, regenerated fresh each time.
• Empty-state page if no .html files found (besides index.html).
• Must work as file:// URL — no server required.
• Do not open individual alignment pages from this skill; individual alignment-producing skills handle their own review pages, and this skill opens or focuses only the central index.

Default Shipping Contract

• No commit/push. Local convenience file, regenerated on demand. Do not commit alignment/index.html unless the user explicitly asks.
• Default next-step routing: After generating the index, suggest contextual next steps based on what was found:
  • If any pages have old modification dates or the scan revealed outdated content: Recommended next command: $upgrade-alignment-pages
  • If the index is freshly built with all pages current: confirm completion with no further routing
  • Do not default to $skills as a generic fallback