Openclaw Refactor Docs logo

Openclaw Refactor Docs

OrganizationPopular
openclaw
openclaw-refactor-docs

Refactor an existing OpenClaw docs page with source-audited preservation, restructuring, and verification.

Overview

Publisheropenclaw
Repositoryopenclaw
Skill nameopenclaw-refactor-docs
Stars
391K
Forks
82.2K
Bundled files
Instructions only
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 openclaw on GitHub. Read the source before you install it.

Installation

Install the Openclaw Refactor Docs 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/openclaw/openclaw/tree/main/.agents/skills/openclaw-refactor-docs
  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/openclaw/openclaw.git /tmp/openclaw
mkdir -p .claude/skills
cp -r /tmp/openclaw/.agents/skills/openclaw-refactor-docs .claude/skills/openclaw-refactor-docs
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Openclaw Refactor Docs 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 Openclaw Refactor Docs 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 Openclaw Refactor Docs 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.

OpenClaw Refactor Docs

Overview

Use this skill when the user gives a target OpenClaw docs page and asks to rewrite, refactor, reorganize, split, shorten, or improve it.

This skill builds on technical-documentation: use that skill for style, page types, structure, examples, discoverability, and verification. This skill adds the rewrite workflow needed to avoid losing accurate behavior during a major docs refactor.

Inputs

Required:

  • A target docs page path, such as docs/plugins/codex-harness.md.

Optional:

  • Desired page type, such as topic page, guide, reference, or troubleshooting.
  • Specific goals, such as shorter main page, move details to reference pages, or align with current CLI behavior.
  • Related source files, schemas, commands, tests, specs, or PRs.

If the target page is missing or ambiguous, ask one concise question before editing. Otherwise, proceed.

Working Contract

Refactor the target page to be more useful, concise, and comprehensive within its stated scope.

Do not treat a rewrite as permission to discard behavior facts. Preserve, verify, move, or explicitly retire existing material. Incorrect docs are worse than verbose docs.

Prefer this split:

  • Topic or guide pages cover the 80/20 path, decisions readers must make, safe setup, smallest reliable verification, common failures, and links onward.
  • Reference pages cover exhaustive fields, defaults, enums, limits, precedence rules, API contracts, narrow internals, and rare debugging details.
  • Troubleshooting pages start from observable symptoms and map to checks, causes, and fixes.

Workflow

1. Load the doc standard

Read ../technical-documentation/SKILL.md first. Apply its page-type, style, examples, navigation, and verification guidance throughout the refactor.

Run pnpm docs:list when available, then read only the target page and the likely entry points, references, or related pages needed for the refactor.

2. Classify the page

Before editing, decide the intended page type from technical-documentation.

If the current page mixes page types, choose the main page type and plan where the other material belongs:

  • Move exhaustive contracts to an existing or new reference page.
  • Move symptom-driven material to an existing or new troubleshooting page.
  • Move narrow setup workflows to a guide when they interrupt the main path.
  • Keep concise routing, decision, and safety details in the main page when readers need them to complete the workflow.

3. Preserve and audit existing facts

Create a working inventory from the old page before rewriting. Include:

  • Config fields, flags, commands, slash commands, env vars, defaults, enums, nullable values, and constraints.
  • Precedence rules, fallback behavior, caps, limits, rate limits, timeouts, lifecycle states, queueing behavior, and compatibility rules.
  • Auth, permission, approval, sandbox, safety, privacy, and destructive-action behavior.
  • Setup requirements, supported versions, dependencies, operating systems, credentials, and account requirements.
  • Error messages, troubleshooting symptoms, diagnostics, and recovery steps.
  • Examples, expected output, command routing tables, and cross-links.

For each fact, choose one outcome:

  • Keep it in the refactored target page.
  • Move it to a specific existing page.
  • Move it to a specific new page.
  • Delete it because current source proves it is obsolete or out of scope.

Do not infer defaults, permissions, policy, timeout behavior, or safety posture from names or intent. Verify them.

4. Find source of truth

Use the nearest authoritative source for each behavior-sensitive claim:

  • Public schema, plugin manifest, generated config docs, or exported types for config fields.
  • CLI implementation, slash-command handlers, help text, and command tests for commands and flags.
  • Runtime source and tests for lifecycle, queueing, permission, fallback, timeout, and provider behavior.
  • Protocol docs, SDK facades, and contract tests for APIs and plugin surfaces.
  • Existing docs only as secondary evidence unless the target is purely conceptual.

If a page promises a reference, compare its tables against the schema, manifest, CLI help, generated docs, or exported types. Missing public fields, defaults, precedence rules, caps, or side effects are correctness bugs.

5. Plan moved material

When moving detail out of the target page, record the destination before editing:

  • Existing page: name the page and section.
  • New page: choose the page type, slug, title, frontmatter summary, doc-schema-version: 1, and read_when hints.
  • Target page: keep a short summary and link from the point where readers need the deeper detail.

Avoid duplicate truth. If the same contract appears in multiple places, choose one canonical page and link to it.

6. Rewrite

Rewrite in this order:

  1. Make the first screen answer what the reader can do and why this page exists.
  2. Put the recommended path before alternatives.
  3. Keep only decision-making and common operational detail in the main flow.
  4. Move exhaustive tables and rare details to the planned reference pages.
  5. Preserve concise routing tables when they help readers choose commands, config paths, harnesses, plugins, providers, or references.
  6. Add troubleshooting from observable symptoms, not internal guesses.
  7. Link related concepts, guides, references, diagnostics, and adjacent tools.

Add doc-schema-version: 1 to the YAML frontmatter of every docs page that the refactor migrates, creates, or materially rewrites. Apply it only to docs page files, not docs.json, glossary JSON, or other non-page metadata. If a migrated page is generated, update the generator so regeneration preserves the marker instead of hand-editing generated output.

Do not leave placeholders such as "TODO", "TBD", or "see docs" unless the user explicitly asks for a draft.

7. Compare old and new

After editing, compare the old and new page:

  • Confirm all behavior-sensitive facts were kept, moved, or intentionally deleted with source-backed reason.
  • Check that the main page still covers the 80/20 scenario end to end.
  • Check that reference pages remain exhaustive for the scope they claim.
  • Check that links from the target page reach moved details.
  • Check that headings are stable, searchable, and action-oriented.

If the refactor deliberately removes relevant material, say where it went or why it was removed in the final report.

8. Verify

Run the smallest reliable docs checks for the touched surface:

  • pnpm docs:list
  • git diff --check -- <touched-files>
  • Targeted pnpm exec oxfmt --check --threads=1 <touched-files>
  • pnpm docs:check-mdx
  • pnpm docs:check-links
  • pnpm docs:check-i18n-glossary when link text, navigation, labels, or glossary surfaces changed
  • Generated-doc checks when schemas, generated config docs, API docs, or generated baselines are touched

Run commands and examples from the page whenever feasible. If you cannot verify a behavior-sensitive claim, either remove the claim, mark the uncertainty in the work-in-progress report, or ask for the missing source.

Final Report

Report:

  • What changed in the target page.
  • What details moved and their destination pages.
  • What source-of-truth checks backed behavior-sensitive claims.
  • What validation ran and what failed for unrelated reasons.

Do not include a long rewrite diary. Lead with remaining risks only if there are any.

Frequently asked questions

What does the Openclaw Refactor Docs AI skill do?

Refactor an existing OpenClaw docs page with source-audited preservation, restructuring, and verification.

Why use Openclaw Refactor Docs on TypingMind?

Because you install it once and use it with any model. Openclaw Refactor Docs 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 Openclaw Refactor Docs in TypingMind?

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

Which AI models can use Openclaw Refactor Docs?

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 Openclaw Refactor Docs?

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

Is the Openclaw Refactor Docs AI skill free?

It is published on GitHub by openclaw. Check the repository for licensing terms. 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 👇