build-skills-index

name	build-skills-index
description	Scan SKILL.md files in .gaai/core/skills/ and .gaai/project/skills/, extract YAML frontmatter, and regenerate separate skills indices for each layer. Core index ships with the OSS framework; project index is project-specific.
license	MIT
compatibility	Works with any filesystem-based AI coding agent
metadata	{"author":"gaai-framework","version":"2.0","category":"cross","track":"cross-cutting","id":"SKILL-CRS-017","tags":["governance","index","discoverability"],"updated_at":"2026-03-17T00:00:00.000Z","status":"stable"}
inputs	[".gaai/core/skills/**/SKILL.md (core framework skills)",".gaai/project/skills/**/SKILL.md (project-specific skills, if present)"]
outputs	[".gaai/core/skills/skills-index.yaml (core skills only — ships with OSS)",".gaai/project/skills/skills-index.yaml (project skills only — if project dir exists)"]

Build Skills Index

Purpose / When to Activate

Activate when:

• A new skill is created (invoked automatically by create-skill Step 6)
• A skill's frontmatter is modified (description, tags, category, id)
• A skill is removed or deprecated
• Either skills-index.yaml is absent or suspected stale (e.g. after a git pull or framework update)

This skill generates the derived catalogs used by agents for fast skill discovery. It does not replace frontmatter — it aggregates it.

skills-index.yaml files are caches. Frontmatter is the source of truth.

---

Process

Step 1 — Scan both skill directories

Scan .gaai/core/skills/ recursively. Collect every file named SKILL.md. If .gaai/project/skills/ exists, scan it too. Ignore non-SKILL.md files. Ignore README.*, skills-index.yaml.

Step 2 — Extract frontmatter from each SKILL.md

For each SKILL.md found, read the YAML frontmatter block (between --- delimiters). Extract the following fields:

Field	Source	Notes
name	frontmatter name	Required
description	frontmatter description	Required
id	frontmatter id or metadata.id	Required
category	frontmatter category or metadata.category	Required
track	frontmatter track or metadata.track	Required
tags	frontmatter tags or metadata.tags	May be absent — default to []
updated_at	frontmatter updated_at or metadata.updated_at	May be absent
path	derived — relative path from .gaai/ to SKILL.md	e.g. core/skills/cross/memory-ingest/SKILL.md

If a required field is missing, log a warning inline in the entry but do not skip — include for visibility. Log any duplicate id values across both layers.

Step 3 — Group entries by track

Organize entries into three groups: discovery, delivery, cross. Within each group, sort alphabetically by name.

Step 4 — Write separate index files

Core index → .gaai/core/skills/skills-index.yaml (core skills only):

# GAAI Core Skills Index
# Source of truth: .gaai/core/skills/*/SKILL.md — core framework skills only
# Regenerate: invoke build-skills-index skill
generated_at: YYYY-MM-DD
total: N

discovery:
  - id: SKILL-DSC-001
    name: create-prd
    source: core
    description: "..."
    category: discovery
    track: discovery
    tags: []
    updated_at: YYYY-MM-DD
    path: core/skills/discovery/create-prd/SKILL.md

Project index → .gaai/project/skills/skills-index.yaml (project skills only, if project dir exists):

# GAAI Project Skills Index
# Source of truth: .gaai/project/skills/*/SKILL.md — project-specific skills only
# Regenerate: invoke build-skills-index skill
generated_at: YYYY-MM-DD
total: N

discovery:
  - id: SKILL-CNT-011
    name: content-plan
    source: project
    description: "..."
    # ... fields

Agents read both indices to get the full skill catalogue.

Step 5 — Report

Return to the invoking agent:

• Core skills count + project skills count
• Any entries with missing required fields (names + fields)
• Any duplicate IDs across both layers
• Confirmation both files were written

---

Automation

The check-and-update-skills-index.js script performs the same work automatically via the post-commit Git hook. This skill exists for manual invocation when the hook doesn't trigger (fresh clone, worktree, CI).

---

Quality Checks

• Every SKILL.md in both directories is represented in the correct index (no silently skipped files)
• No manually written entries — all entries derived from frontmatter
• Duplicate id values flagged across both layers (do not silently deduplicate)
• generated_at reflects the date of this run
• Core index contains zero source: project entries
• Project index contains zero source: core entries

---

Non-Goals

This skill must NOT:

• Edit any SKILL.md file
• Make decisions about which skills are valid or relevant
• Merge duplicate skills or resolve conflicts — only report them
• Be invoked as a dependency of other skills (only agents call this)

This skill reads and aggregates — it does not evaluate or decide.