skills.rest
umairshahid622
umairshahid622COMMUNITY
Explore all skills by this author →
GitHub

skill-developer

Create and configure Claude Code skills with trigger rules and hook-based activation.

Updated Sep 1, 2026
Search Tags:#automation#skill development#claude code#hooks#regex#trigger patterns#yaml frontmatter
One-click install
npx skills add https://github.com/umairshahid622/perfumora --skill skill-developer-umairshahid622
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: skill-developer
Source: https://github.com/umairshahid622/perfumora/tree/main/perfumora-admin/.claude/skills/skill-developer
Command: npx skills add https://github.com/umairshahid622/perfumora --skill skill-developer-umairshahid622

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Building effective Claude Code skills requires understanding YAML frontmatter, trigger patterns, hook mechanics, and configuration schemas, and mistakes cause skills to never activate or fire incorrectly. ## Core Features & Use Cases - Skill Creation Workflow: Step-by-step process for authoring SKILL.md files with proper frontmatter, following the 500-line rule and progressive disclosure pattern. - Trigger Configuration: Complete reference for keywords, intent regex patterns, file path globs, and content patterns in skill-rules.json, plus a copy-paste pattern library. - Hook System Documentation: Deep dives into UserPromptSubmit and PreToolUse hook flows, exit code behavior, session state tracking, and skip conditions. - Use Case: When you need to add a guardrail skill that blocks Prisma edits until database column names are verified, this Skill provides the exact JSON schema, block message template, and testing commands to implement it correctly. ## Quick Start Ask the AI to create a new Claude Code skill with trigger rules for your chosen topic and test the activation hooks.

Frequently Asked Questions about skill-developer

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

FAQPage Schema
How do I create a new Claude Code skill?▼

Create a SKILL.md file with YAML frontmatter containing name and description in .claude/skills/{skill-name}/, then add a matching entry to skill-rules.json with trigger keywords and intent patterns. Test activation by piping a sample prompt to the skill-activation-prompt hook with npx tsx.

What trigger types can activate a Claude Code skill?▼

Skills support four trigger types: keyword substring matching, intent regex patterns on prompts, file path glob patterns, and content regex patterns matched against file contents. Keywords and intent patterns drive UserPromptSubmit suggestions, while file triggers drive PreToolUse enforcement.

Why is my skill not triggering on user prompts?▼

Common causes include keywords missing from the prompt, intent regex patterns that are too specific, a name mismatch between SKILL.md and skill-rules.json, or invalid JSON syntax. Validate the JSON with jq and test the hook manually with a sample prompt.

What is the difference between block, suggest, and warn enforcement?▼

Block uses PreToolUse with exit code 2 to prevent Edit or Write execution until the skill is used, reserved for critical guardrails. Suggest injects an advisory reminder via UserPromptSubmit without blocking. Warn is a low-priority advisory level that is rarely used.

How do I stop a guardrail skill from blocking repeatedly?▼

Session tracking automatically skips blocking after the skill is used once per session via state files. You can also add a // @skip-validation marker comment to a file or set an environment variable like SKIP_SKILL_GUARDRAILS to disable enforcement.

What is the 500-line rule for SKILL.md files?▼

Anthropic best practice recommends keeping SKILL.md under 500 lines so it loads efficiently into context. Detailed content should move into reference files with tables of contents, following the progressive disclosure pattern.

Related Skills

Explore More Software Engineering Skills→
cescraflicescrafli

security-scan

Scan .claude directory configurations for security vulnerabilities and misconfigurations.

Community
Intermediate
fallow-rsfallow-rs
4.4k

conformance-loop

Compares Fallow analysis output against competing tools and verified source truth to improve accuracy.

Official
Intermediate
ddttomddttom
2

skill-developer

Automate Claude Code skill governance and auto-activation rule management.

Community
Intermediate
Sir-chawakornSir-chawakorn
9

setup-lint-format-precommit

Configures linting and formatting tools with pre-commit hooks and CI enforcement.

Community
Advanced