skills.rest
odysseyalive
odysseyaliveCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

skill-builder

Audit and optimize Claude Code skill files across repositories.

2|1|Updated Feb 8, 2017
Search Tags:#audit#hooks#agents#enforcement#optimize#claude-code#skill-builder
One-click install
npx skills add https://github.com/odysseyalive/dotfiles --skill skill-builder-odysseyalive
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: skill-builder
Source: https://github.com/odysseyalive/dotfiles/tree/main/.claude/skills/skill-builder
Command: npx skills add https://github.com/odysseyalive/dotfiles --skill skill-builder-odysseyalive

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Skill Builder automates the discovery, auditing, optimization, and enforcement of Claude Code skills so teams can maintain consistent, drift-resistant skill repositories without manual review. It centralizes checks for frontmatter, grounding, hooks, agents, and CLAUDE.md extraction candidates to prevent broken behavior and contextual drift across long conversations.

Core Features & Use Cases

  • Project-wide Audits: Run a full, read-only audit that aggregates optimize, agents, and hooks findings and ranks priority fixes.
  • Optimize & Refactor: Propose and (with execute mode) apply safe restructures that move references, preserve directives verbatim, and split large reference files into enforcement boundaries.
  • Agent & Hook Recommendations: Detect where context: none agents or PreToolUse hooks are required, produce agent templates, and wire hook scripts.
  • Bootstrap & CLAUDE.md Extraction: Extract domain content from CLAUDE.md into new skills for first-run or migration scenarios.
  • Self-Heal & Ledger Integration: Install companion skills (self-heal, awareness-ledger) and seed institutional memory; suggest capture opportunities and embed surgical repair observers.
  • Execute Mode Discipline: When executing changes, generate explicit TaskCreate task lists and update progress via TaskUpdate to maintain resumable, auditable operations.
  • Use Case Example: Quickly audit a newly cloned repo, get prioritized fixes (frontmatter, hooks, agent gaps), and apply selected changes with a single, resumable task plan.

Quick Start

Run /skill-builder audit to produce a display-mode system-wide audit report that lists optimization, agent, and hook recommendations and offers executable task plans.

Frequently Asked Questions about skill-builder

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I audit Claude Code skills across my repository?▼

Auditing Claude Code skills involves running a read-only system-wide scan that aggregates optimization, agent, and hook findings to rank priority fixes across .claude/skills directories. The output is a display-mode report offering executable task plans.

What is the best way to enforce frontmatter and grounding in Claude skills?▼

Enforcing frontmatter and grounding in Claude skills is achieved by applying execution discipline requirements that validate file structure during audits. This prevents broken behavior and contextual drift across long conversations by ensuring all skill files meet structural standards.

How do I extract domain content from CLAUDE.md into new skills?▼

Extracting domain content from CLAUDE.md into new skills uses a bootstrap process designed for first-run or migration scenarios. This feature identifies extraction candidates within the file and generates properly structured skill files in the .claude/skills directory.

Does skill-builder work with PreToolUse hooks and context none agents?▼

Yes, it detects where PreToolUse hooks or context: none agents are required within your repository. It produces agent templates and wires hook scripts to enforce boundaries, ensuring safe execution and preventing unauthorized tool usage during long conversations.

Can I apply refactoring changes to skill files automatically?▼

You can apply refactoring changes automatically using execute mode, which generates explicit TaskCreate task lists and updates progress via TaskUpdate. This maintains resumable, auditable operations while safely moving references and splitting large files into enforcement boundaries.

When do I need self-heal and awareness-ledger integration for Claude skills?▼

Self-heal and awareness-ledger integration is needed when you want to seed institutional memory and embed surgical repair observers in your skills. This companion integration suggests capture opportunities to prevent contextual drift and maintain consistent repository behavior over time.

Related Skills

Explore More Software Engineering Skills→
RBozydarRBozydar

create-agent-skills

Create, audit, and package Claude Skills with SKILL.md structure.

Community
Advanced
imgompandaimgompanda
139

fireauto-team-guide

Coordinate multi-agent teams on a shared task board.

Community
Advanced
UniquecreteUniquecrete

supermemory

Stores and retrieves long-term agent memory via Supermemory's knowledge-graph API.

Community
Intermediate
Sharper-FlowSharper-Flow
4

adv-arch-detection

Detect architectural inconsistencies and dependency violations across technology stacks.

Official
Advanced