question-answer
Skill1 starsChanged 3 months agoMITView on GitHub
Use when the user wants a read-only explanation of existing code behavior, architecture, control flow, data flow, dependencies, or implementation decisions and has not requested project changes.
In this repository13 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- bug-analysisSkill
- cavemanSkill
- code-review-with-filesSkill
- diagnoseSkill
- fetch-yuque-docSkill
- git-commit-workflowSkill
- handoffSkill
- improve-codebase-architectureSkill
- ios-i18n-workflowSkill
- java-to-nextjsSkill
- java-to-swiftSkill
- reconstructing-ui-from-screenshotsSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 15
Use this skill for requests to explain:
- Distinctive: line 36
Use external documentation only when the answer depends on an external API, version, or current behavior.
- Distinctive: line 37
If evidence is insufficient, state what is unknown and what would need inspection.
- Distinctive: line 55
Read-only: Do not modify files or mutate state while this skill applies
- Distinctive: line 61
Reading the whole repository: Start with the narrowest relevant path and expand only when evidence requires it.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Codebase Question Answering
Overview
Answer codebase questions from evidence rather than assumptions. Stay read-only while this skill applies, respond in the user's language, and match the requested depth and format.
Scope
Use this skill for requests to explain:
- Code behavior, control flow, or data flow
- Architecture, dependencies, or implementation decisions
- How an existing feature or subsystem works
This skill does not apply when any requested outcome includes:
- Editing, adding, deleting, or generating project files
- Implementing a feature or fix
- Running commands that mutate project or external state
- Reviewing a change for defects
- Diagnosing an uncertain bug, crash, or regression
For a mixed request such as “explain this and fix it,” use the appropriate change workflow and satisfy both outcomes. Do not let this skill's read-only constraint override an explicit change request.
Workflow
- Identify the exact question and requested level of detail.
- Inspect the smallest relevant code slice: definitions, callers, configuration, tests, and dependencies as needed.
- Trace the behavior from concrete evidence. Separate observed facts from inference.
- Use external documentation only when the answer depends on an external API, version, or current behavior.
- If evidence is insufficient, state what is unknown and what would need inspection; do not guess.
Answer Contract
Shape the response in this order:
- Direct answer or conclusion
- Relevant code evidence, including file and symbol references when available
- Reasoning that connects the evidence to the conclusion
- Caveats, trade-offs, or next steps only when they help answer the question
Honor explicit length and format requests. A simple question may need one sentence; use sections only when they improve clarity.
Quick Reference
- Facts: Supported directly by inspected code, configuration, tests, or documentation
- Inference: Label it and explain the evidence behind it
- Unknown: State the missing evidence instead of presenting a guess as fact
- Read-only: Do not modify files or mutate state while this skill applies
Common Mistakes
- Using this skill for a mixed change request: Switch to the appropriate change workflow.
- Restating before answering: Lead with the answer unless clarification is required.
- Reading the whole repository: Start with the narrowest relevant path and expand only when evidence requires it.
- Replacing repository evidence with generic best practices: Explain the current code first; add external guidance only when relevant.
- Forcing a fixed template: Match the response shape to the question.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

