lararosekelley/dotfiles/teach

Claude Code and agent skill by lararosekelley/dotfiles

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Teach a repository-specific concept or change through a source-backed mental model and worked example. Use when the user invokes /teach or explicitly requests a deeper explanation to understand code. Do not trigger for a brief factual answer, ordinary code review, or an implementation request.

Install

npx vskill@latest install lararosekelley/dotfiles/teach

SKILL.md

Teach

Build understanding of the code without overwhelming the reader. Follow current repository instructions, use plain language, and keep the work read-only.

Help dispatch

If the entire argument is help, --help, or -h, print Usage and stop before investigating. Otherwise read the topic and requested level from the invocation or forwarded arguments.

Usage

/teach [topic or change|help] [level or focus]

Explain a concept with a small mental model, one concrete example, and evidence from the actual code. With no topic, use a clear topic from this conversation or ask one question. Default to an experienced developer new to this subsystem.

text
/teach how this PR changes retries
/teach sync filtering, focus on why both directions use the repo
/teach worktrees step by step, assume I know branches
/teach help

Add a level, focus, or "step by step" in ordinary language. In step-by-step mode, show one concept and wait; next continues and stop returns to normal conversation. Otherwise give a compact explanation without a mandatory quiz. Use /how for a short execution trace and /why for a historical-rationale investigation.

Build the explanation

  1. Identify what the user wants to understand and what they already know from

the conversation. Do not force an interview when a reasonable default suffices.

  1. Read the relevant code and trace one representative input to its result.

Reuse evidence already gathered, verifying it against current files when needed. Do not automatically run /how and /why as separate full investigations.

  1. Investigate history only when it helps answer the requested rationale. Separate

documented intent from a useful explanation of the current design. Cite sources; do not turn an analogy into a claim about how the implementation works.

  1. Choose one mental model and a small worked example. Distinguish illustrative

inputs from observed runtime results. Include one important boundary or failure case when it sharpens understanding. For a change, contrast before and after.

  1. Stay read-only and follow the user's delegation policy. Do not modify code,

start services, or run side-effectful demonstrations without a separate request.

Present and continue

For a focused topic, aim for roughly 200-400 words initially: the central idea, one example, and the key constraint. Use short code snippets or a small ASCII diagram only if they reduce explanation. Cite the few source locations that let the user check the explanation; do not list every file you opened.

Honor the requested level and expand when needed rather than enforcing a word budget at the expense of clarity. Do not require a quiz, ask repeated comprehension questions, or offer a long menu of next lessons. If the user requests step-by-step teaching, wait after each concept and accept ordinary-language questions or next.

If evidence is missing, say what cannot be established. Do not present a plausible explanation as proven behavior. On stop, return to normal conversation without another lesson prompt or extra investigation.

Security

  • Security scanPassed, 95 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

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
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/lararosekelley/dotfiles/teach/badge)](https://verified-skill.com/skills/lararosekelley/dotfiles/teach)

Something wrong with this skill or its page? Tell us.