davesnx/setup/how
Claude Code and agent skill by davesnx/setup
- community
Checking that the SKILL.md file is still in the repository.
Explain or research how something works, in this codebase or outside it, by exploring code or primary sources and producing a clear, sourced explanation. Use for "how does X work", "research X", "trace X", or "where is X used" — X can be local code or an external topic, such as a library, API, other repo, or knowledge base. Optionally critique the architecture for issues.
Install
npx vskill@latest install davesnx/setup/howSKILL.md
How
Explain enough of the system for the reader to follow its behavior and find the relevant code. For a question about an external topic, explain it from primary sources instead, with a source on each claim. This is read-only work unless the user separately asks for edits.
Trace the system
- Define the scope from the question. Resolve factual uncertainty from the code;
ask only when different interpretations would change the work.
- Follow entry points, callers, state ownership, and failure paths across the
relevant modules. Separate observed behavior from intended design.
- Stop when the important path is accounted for, or name the missing evidence.
Explore directly by default. Use parallel read-only explorers only for separate parts that justify the extra work. Give each a distinct question and use references/explorer-prompt.md when delegating. Let the host configuration select models. Synthesize the findings yourself and verify disagreements in the code.
Research outside the codebase
Some questions point outside this repository: a third-party library, an API, another repo, or a knowledge base.
- For a remote git repo, use the librarian skill for a local cached checkout,
then trace it like local code.
- For docs, APIs, specs, and knowledge bases, read the primary source itself,
not a secondary write-up of it. Follow each claim back to the page or file that states it.
- Give each claim a source: a URL, a file path, or a doc section. A caller
that records findings on a ticket, such as wayfinder, needs that source on each line it keeps.
- Stop when the claims the question needs are covered, or name what you could
not verify.
Explain
Lead with the answer. Describe the flow and the reasons for its boundaries, with concrete file and symbol references. Add a package map, definitions, diagram, or cautions only when they help the reader. Match the requested length; a fixed set of headings is not required.
Critique when requested
Use references/critique-rubric.md to inspect correctness, coupling, change cost, and testability. Verify each concern before presenting it. Distinguish changes worth making now, tradeoffs to consider, and claims rejected by the evidence.
For a broad or risky critique, or an explicit request for independent reviews, give reviewers distinct concerns using references/critic-prompt.md. Keep the review read-only. Report the explanation and supported findings, not a transcript of the review process.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingagents/skills/how/SKILL.md
Popularity
- Repository stars
- 11
- Repository forks
- 1
- 7-day trend score
- 0
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/davesnx/setup/how)