skill-builder
Skill Builder
Build or revise Claude skills from evidence, not instinct.
Core rule
Inspect the current workspace first.
For new skills, major redesigns, or multi-source merges, research similar remote skills before drafting.
Show skills.markdown and a comparison table before you synthesize the final result.
Default workflow
-
Classify the job before drafting.
- Decide whether this is new-skill creation, substantial redesign, or a small local cleanup.
- Use the lightest workflow that still preserves evidence quality.
-
Inspect the local workspace first.
- Run
treeon the current working directory, or produce an equivalent tree-style listing iftreeis unavailable.
- Run
More from yigitkonur/skills-by-yigitkonur
run-research
Use skill if you are answering one technical research question with current web evidence, Reddit practitioner experience, and source-backed synthesis, optionally fanned out across parallel subagents.
42run-agent-browser
Use skill if you are driving the agent-browser CLI for ad hoc browser automation — @ref snapshots, sessions, forms, extraction, screenshots, headed/stealth, or provider runs.
38publish-npm-package
Use skill if you are publishing to npm via GitHub Actions release workflow with trusted publishing, NPM_TOKEN, provenance, semantic-release, changesets, release-please, or fixing npm publish CI.
35init-agent-config
Use skill if you are creating, auditing, or migrating CLAUDE.md, AGENTS.md, or REVIEW.md instruction files, folder-scoped agent guidance, or cross-agent companion entrypoints.
34build-mcp-use-server
Use skill if you are building TypeScript MCP servers with mcp-use/server — server.tool, response helpers, ctx.auth, sessions, transports, widgets, Inspector, deploy.
34build-skills
Use skill if you are creating or substantially revising a Claude skill and need workspace-first evidence, remote comparison, and repo-fit synthesis before writing SKILL.md.
32