Help us improve
Share bugs, ideas, or general feedback.
From synapse-a2a
Design, review, and improve Claude/Codex skills based on Anthropic's "The Complete Guide to Building Skills for Claude". Use when creating a new skill, rewriting SKILL.md frontmatter and workflows, fixing under-triggering or over-triggering, designing scripts/references/assets, building test cases, or preparing a skill for upload/distribution.
npx claudepluginhub s-hiraoku/synapse-a2a --plugin synapse-a2aHow this skill is triggered — by the user, by Claude, or both
Slash command
/synapse-a2a:anthropic-skill-creatorThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Build skills that trigger correctly, execute reliably, and stay concise.
Step-by-step walkthrough for designing, writing, testing, and maintaining Claude Skills, including how to write trigger descriptions and structure SKILL.md files.
Guides creation, improvement, auditing, testing, and distribution of Claude Code skills using Anthropic's official methodology.
Guides developers in creating and updating effective Claude skills with best practices, core principles, resource links, and workflow advice for specialized extensions.
Share bugs, ideas, or general feedback.
Build skills that trigger correctly, execute reliably, and stay concise.
name, description) for trigger selection.SKILL.md body for core workflow.references/ for deep details loaded only when needed.scripts/, references/, assets/).Capture for each use case:
If use cases are vague, ask targeted follow-up questions before authoring.
Write only what is needed to trigger correctly.
Required fields:
name: kebab-case, matches folder namedescription: include BOTH
Description formula:
[What it does] + [When to use it] + [Key capabilities]Good pattern:
Bad pattern:
Precision controls:
scripts/: deterministic checks/transforms used repeatedlyreferences/: large docs, domain logic, variant-specific guidanceassets/: templates and artifacts used in outputsRules:
SKILL.md compact; link to references/.README.md, changelog, etc.).Write imperative, testable steps.
Include:
For critical checks, prefer scripts over prose-only checks.
Run these three test groups.
Target benchmark:
Validate:
Compare with vs without skill:
Under-triggering signals:
Fix:
description.Over-triggering signals:
Fix:
Execution issues:
Fix:
SKILL.md: file must be exactly SKILL.md.--- and valid syntax.references/; reduce enabled skills.Scaffold a new skill directory:
# Use default path (.agents/skills)
bash scripts/new_skill.sh <skill-name>
# Place in .claude/skills directory
bash scripts/new_skill.sh <skill-name> --path .claude/skills
references/checklist.mdreferences/prompt-templates.mdreferences/patterns.mdreferences/source.md: source PDF and extracted principlesreferences/trigger-examples.md: trigger/non-trigger examples for description tuningreferences/failure-remedies.md: failure patterns and minimum effective fixes