Code Walkthrough logo

Code Walkthrough

Organization
skillmds
code-walkthrough

Guide a visible, verified code walkthrough in Neovim with Terminal Control and Navi, optionally recording the presentation with OBS or retaining terminal evidence. Use when the user wants an agent-led code tour, code-along, live source explanation, or recorded terminal walkthrough.

Overview

Publisherskillmds
Repositoryskillmd
Skill namecode-walkthrough
Stars
249
Forks
3
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by skillmds on GitHub. Read the source before you install it.

Installation

Install the Code Walkthrough AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/skillmds/skillmd/tree/main/plugins/agents-mcp/skills/code-walkthrough
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/skillmds/skillmd.git /tmp/skillmd
mkdir -p .claude/skills
cp -r /tmp/skillmd/plugins/agents-mcp/skills/code-walkthrough .claude/skills/code-walkthrough
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Code Walkthrough in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Code Walkthrough on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Code Walkthrough is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Code Walkthrough

Use a verified tour: establish the real code and output first, then guide attention through one immutable Navi tour.

Prepare The Session

  1. Confirm termctrl, Neovim 0.10 or newer, and navi.nvim are available.

  2. Start Neovim in the project root as a visible foreground session:

    bash
    termctrl run walkthrough -- nvim
  3. Use Terminal Control from another process to inspect and drive that same session. Prefer interact through the Terminal Control MCP when available; otherwise use termctrl show, termctrl send, and termctrl wait.

  4. Resize before composing the tour if the visible terminal is too narrow. Keep the size stable while presenting or recording.

The session is ready when the human and agent see the same settled Neovim screen and the agent can inspect it through Terminal Control.

Establish The Journey

  1. Ask what the viewer should understand or be able to explain afterward when the outcome is unclear.
  2. Read the relevant code and run the narrowest command that proves its current behavior.
  3. Inspect the actual visible result. Do not write notes from an expected or remembered result.
  4. Choose one conceptual journey. A useful journey often moves from public behavior to the mechanism that produces it, then to the observed result.

Do not combine an overview, failure diagnosis, implementation tour, and fix into one sequence. The journey is established when each planned stop is necessary to answer one coherent question.

Author One Immutable Tour

Create a JSON file outside the source tree when the tour is temporary. Use literal patterns when they are unique and stable; use line numbers only when exact positions are the point.

json
[
  {
    "file": "src/example.test.ts",
    "pattern": "it(\"updates subscribers\"",
    "end_pattern": "expect(runs).toBe(2)",
    "message": "The public contract starts here: one write must cause one additional run."
  },
  {
    "file": "src/example.ts",
    "pattern": "export function notify",
    "end_pattern": "subscriber()",
    "message": "The implementation reaches every subscriber through this loop."
  }
]

Load it without modifying it:

vim
:NaviLoad /absolute/path/to/tour.json

Apply these rules:

  • Create at most one tour per assistant response.
  • Treat a shown tour as immutable. Do not rewrite its stops or notes in the same response.
  • Start at the behavior or call site before entering implementation details.
  • Use one range stop for adjacent lines that form one simple mechanism.
  • Use multiple stops only for a genuine journey between distinct locations.
  • Put most of the explanation in short, casual Navi notes; keep chat brief.
  • Keep concise verified output at the decisive assertion or result when it helps explain the behavior.
  • If verification disproves the explanation, stop and say so. Clear or replace the tour only in the next response.

The tour is complete when every stop resolves, the first stop is visible, and its notes agree with the verified code and output.

Present The Tour

Let the viewer control the pace unless asked to drive:

text
]n           next stop
[n           previous stop
<leader>np  pick a stop
<leader>nc  clear the tour

These are recommended mappings, not Navi defaults. If they are unavailable, use :NaviNext, :NaviPrev, :NaviPick, and :NaviClear.

Pause for questions without replacing the active tour. A follow-up explanation may focus the current range or inspect output, but a different conceptual journey belongs in a later tour.

Record With OBS

Use OBS when the desired artifact is the human-facing presentation, including natural pacing, narration, window chrome, or other visual context.

  1. Add a Window Capture source for the terminal window containing termctrl run.
  2. Crop unrelated tabs, prompts, notifications, and private paths before recording.
  3. Set the terminal size, font size, OBS canvas, and capture crop before authoring the tour so notes wrap exactly as they will in the recording.
  4. Perform a short test recording. Verify text legibility, microphone level, terminal contrast, and that Navi notes fit without clipping.
  5. Start OBS recording, present the immutable tour, then stop recording after the final frame has held long enough to read.

OBS is the primary recording in this branch. Terminal Control remains the agent-control and verification surface.

Retain Terminal Evidence

Use Terminal Control artifacts when the user wants a terminal-native replay, a screenshot, or debugging evidence rather than a full OBS presentation.

  • Capture the current screen at any time with termctrl save walkthrough --format png --out artifacts/current.

  • Recording must be enabled when the session starts; it cannot currently be added to an already-running foreground session.

  • For a terminal-native recording, restart deliberately with --record before presenting:

    bash
    termctrl run walkthrough --record artifacts/walkthrough.termctrl -- nvim
  • Add markers at meaningful moments, then inspect or export the recording after the session ends:

    bash
    termctrl mark walkthrough contract
    termctrl mark walkthrough mechanism
    termctrl markers artifacts/walkthrough.termctrl
    termctrl video artifacts/walkthrough.termctrl --out artifacts/walkthrough.mp4

Do not record or retain terminal input, source, paths, or output that may contain secrets unless the user explicitly needs that artifact.

Finish

Clear the tour and stop the session unless the user wants it left open:

vim
:NaviClear
bash
termctrl stop walkthrough

Report the verified behavior, the journey presented, and any retained artifact paths. The walkthrough is complete when the viewer has traversed the journey, the terminal session has the requested final state, and every promised artifact has been checked.

Tool References

Frequently asked questions

What does the Code Walkthrough AI skill do?

Guide a visible, verified code walkthrough in Neovim with Terminal Control and Navi, optionally recording the presentation with OBS or retaining terminal evidence. Use when the user wants an agent-led code tour, code-along, live source explanation, or recorded terminal walkthrough.

Why use Code Walkthrough on TypingMind?

Because you install it once and use it with any model. Code Walkthrough is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Code Walkthrough in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/skillmds/skillmd/tree/main/plugins/agents-mcp/skills/code-walkthrough. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Code Walkthrough?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Code Walkthrough?

As many as you like. As long as a model supports skills, you can use Code Walkthrough with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Code Walkthrough AI skill free?

Yes. It is published on GitHub by skillmds under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇