digipulse-engineering

build-skills-index

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.

digipulse-engineering 163 26 Updated 7mo ago
GitHub

Install

npx skillscat add digipulse-engineering/gaai-framework/build-skills-index

Install via the SkillsCat registry.

SKILL.md

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.cjs 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.