agent-browser
Browser automation CLI for AI agents. Use when the user needs to interact with websites,…
Audit code quality and architectural health or plan refactoring using Fowler code smells and techniques; write prioritized findings to docs/_refacs/. Excludes formatting, performance, and security audits.
$ npx -y skills add pedronauck/skills --skill refactoring-analysis --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/refactoring-analysisContext preview
The summary Claude sees to decide when to auto-load this skill.
Audit code quality and architectural health or plan refactoring using Fowler code smells and techniques; write prioritized findings to docs/_refacs/. Excludes formatting, performance, and security audits.
name: refactoring-analysis description: "Audit code quality and architectural health or plan refactoring using Fowler code smells and techniques; write prioritized findings to docs/_refacs/. Excludes formatting, performance, and security audits." disable-model-invocation: true metadata: author: Pedro Nauck github: https://github.com/pedronauck repository: https://github.com/pedronauck/skills
Find changes that reduce demonstrated maintenance cost while preserving behavior. Name the code smell, source evidence, and concrete improvement; a textbook category without a practical consequence is not a finding.
1. Resolve the scope from the request and active context. For a broad request, map the codebase and prioritize dependencies/hotspots; do not stop for confirmation merely because it exceeds 50 files. Disclose sampled or uninspected areas. 2. Read the relevant sections of `references/code-smells-catalog.md` and `references/refactoring-techniques.md` for the candidates found. Catalog thresholds are signals, not mandatory edits or a requirement to produce findings in every category. 3. Trace ownership, coupling, call sites, and public boundaries. Consult `references/solid-ddd-context.md` only for a domain-model question; ordinary code does not need a SOLID pass. 4. Rank confirmed opportunities by user/maintenance impact, frequency, risk, and effort. For the highest-impact items, describe the before/after behavior and owning verification. Reuse existing coverage; missing tests do not automatically require a new suite at every layer. 5. A requested report defaults to `docs/_refacs/<YYYYMMDD>-<slug>.md`. Adapt `assets/refactoring-report-template.md`; a narrow request can use a concise response. Check material citations and reasoning with `checklists/analysis-checklist.md` as applicable. 6. If the user requested an audit only, report findings. If they already authorized refactoring, continue the selected work without asking again. Preserve published compatibility and unrelated changes.
A curated collection of 87 agent skills for Claude Code and compatible AI coding assistants — 40 original (⭐️), 34 hand-picked (💎), and 13 marketing & business (📣).
Repo: pedronauck/skills
Browser automation CLI for AI agents. Use when the user needs to interact with websites,…
Answer questions about the AI SDK and help build AI-powered features. Use when developers:…
Autonomously optimize any Claude Code skill by running it repeatedly, scoring outputs against…
Centrifugo real-time messaging server expert for WebSocket PUB/SUB, channel management, JWT…
Retrieves up-to-date documentation, API references, and code examples for any developer…
Diátaxis Documentation Expert. An expert technical writer specializing in creating…