Create Claude Code skills with SKILL.md frontmatter and hook tests.
npx skills add https://github.com/adteco/glapi --skill skill-developer-adteco
Please help me install this Agent Skill. Skill: skill-developer Source: https://github.com/adteco/glapi/tree/main/.claude/skills/skill-developer Command: npx skills add https://github.com/adteco/glapi --skill skill-developer-adteco
Skill-developer brings order to Claude Code skill creation so teams stop missing trigger keywords, exceed the 500-line rule, or forget progressive disclosure reference files when defining new guardrails or domain guidance.
Use this guide to craft new SKILL.md files, align them with skill-rules.json, and run the prescribed hook tests before committing.
High-intent search queries and answers about installing and using this skill.
To create SKILL.md files for Claude Code, include concise frontmatter, document trigger patterns, sync names with skill-rules.json, and use progressive disclosure reference files for detailed knowledge. Keep content under 500 lines while pointing to external references for deeper context.
Claude Code hook behaviors orchestrate UserPromptSubmit and PreToolUse flows using exit codes, session tracking, and skip markers. PreToolUse scripts intercept tool execution to enforce guardrails, returning specific exit codes that control whether the tool call proceeds or gets blocked.
Configure trigger patterns by documenting intent keywords in SKILL.md and syncing them with skill-rules.json. Define guardrails using PATTERNS_LIBRARY and HOOK_MECHANISMS reference files, then test PreToolUse scripts to verify enforcement messages match SKILL_RULES_REFERENCE specifications.
Guardrails fail to trigger when SKILL.md keywords are missing or misaligned with skill-rules.json entries. Verify your trigger patterns match user prompts, check hook exit-code behavior in PreToolUse scripts, and confirm skip markers aren't inadvertently bypassing the enforcement logic.
Progressive disclosure in skill development keeps SKILL.md concise by placing detailed knowledge in separate reference files like PATTERNS_LIBRARY and HOOK_MECHANISMS. This approach satisfies the 500-line limit while preserving deep domain guidance for complex guardrail and hook configurations.
Test Claude Code hook configurations by running prescribed hook tests that validate UserPromptSubmit and PreToolUse flows, verify exit-code behavior, and confirm guardrail messages match SKILL_RULES_REFERENCE. Check session tracking and skip markers to ensure reliable enforcement before committing.
ZeroGravitySkin-RonAutomate Claude Code skill creation with SKILL.md authoring and skill-rules.json configuration.
markus41Configure lifecycle event hooks with shell commands, HTTP requests, or LLM prompts.
joshka0Author Claude Code skills with SKILL.md frontmatter and trigger rules.
ddttomAutomate Claude Code skill governance and auto-activation rule management.
