Ensure Docs logo

Ensure Docs

Organization
existential-birds
ensure-docs

Verify documentation coverage and generate missing docs interactively

Overview

Publisherexistential-birds
Repositorybeagle
Skill nameensure-docs
Stars
82
Forks
8
Bundled files
1
LicenseApache-2.0
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.

  • 1 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by existential-birds on GitHub. Read the source before you install it.

Installation

Install the Ensure 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.
  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/existential-birds/beagle.git /tmp/beagle
mkdir -p .claude/skills
cp -r /tmp/beagle/plugins/beagle-docs/skills/ensure-docs .claude/skills/ensure-docs
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Ensure 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 Ensure 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 Ensure 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.

Ensure Documentation Coverage

Verify documentation coverage across a codebase, report gaps, and generate missing docs. If the agent supports subagents, dispatch one verifier per detected language in parallel; otherwise run the same per-language verification sequentially — the output is identical either way.

Coverage has two complementary lenses, and a healthy project needs both:

  1. Symbol coverage — are the functions, classes, and modules documented to the language's standard (docstrings, JSDoc, GoDoc)? This is the per-language verification below.
  2. Diataxis type balance — does the doc set as a whole serve all four user needs: a Tutorial to learn, How-To guides for tasks, Reference for lookups, and Explanation for understanding? A codebase can have 100% docstring coverage and still have no way for a newcomer to get started. See the Diataxis balance check below.

Workflow

Complete steps in order. Do not advance until each step’s Pass is satisfied.

  1. Language detection — Follow Phase 1 (language detection) in references/workflow.md.

    • Pass: For each language you will verify, you have evidence of at least one matching source file (counts or command output); if none qualify, stop with a short “no applicable languages” message and do not run verifiers.
  2. Load standards — Read the sections for your detected languages (language standards, verifier prompts, consolidation format) in the same reference file.

    • Pass: You can state which standard applies per language (e.g. Google docstrings, JSDoc, GoDoc) before verification begins.
  3. Verification — Verify each qualifying language using the verifier prompts and JSON output shape in the reference (Phase 2). If the agent supports subagents, run one verifier per language in parallel; otherwise run them sequentially.

    • Pass: Each completed verification returns parseable JSON including language, files_scanned, and findings (array, possibly empty).
  4. Diataxis balance check — Run the Diataxis type balance check against the project's existing docs (e.g. a docs/ tree, README, or wiki).

    • Pass: You can state, for each of the four types (Tutorial, How-To, Reference, Explanation), whether the project has at least one document serving that need, and you have noted any missing or thin type.
  5. Consolidated report — Merge results per Phase 3 (summary table, severity grouping, detailed findings if requested). Include the Diataxis balance alongside symbol coverage.

    • Pass: The user sees the merged report (inline or written to an agreed path) — covering both symbol coverage and Diataxis type balance — before you claim the audit is done or propose fixes.
  6. Generation — Only if --report-only is not set: offer choices per Phase 4; apply doc edits only after an explicit user choice to generate. For a missing Diataxis type, route generation through draft-docs for the relevant type rather than generating inline.

    • Pass: No documentation edits for gaps until the user selects an option that includes generation; if they decline or choose report-only behavior, end after the report.
  7. Post-edit verification — After any generation, run or offer the linter commands in Phase 5 of the reference for languages you changed, when those tools exist in the repo.

    • Pass: Linter run completed with output captured, or N/A with a one-line reason (e.g. tool not configured); remaining issues are listed or cleared.

Diataxis Type Balance Check

Survey the project's prose documentation (a docs/ tree, README, wiki, or doc site) and classify what exists into the four Diataxis types. Use the compass to classify — action or cognition? acquisition or application? — per docs-style/references/diataxis-compass.md.

Report the balance as a table:

TypePresent?Notes
Tutorial (learning)yes / no / thine.g. "No getting-started / first-project guide"
How-To (tasks)yes / no / thine.g. "Several task guides under docs/how-to/"
Reference (lookup)yes / no / thine.g. "API reference generated, but no CLI reference"
Explanation (understanding)yes / no / thine.g. "No architecture / design-rationale docs"

Flag, in priority order:

  1. A missing type — the doc set serves none of that user need. The most common and most damaging gap is a missing Tutorial: a project can have exhaustive reference and still leave a newcomer with no way in.
  2. A thin type — present but doesn't cover the project's major features or surfaces.
  3. Mixed documents — a single page trying to be two types at once (e.g. reference tables embedded in a how-to). Recommend splitting via improve-doc.

Do not propose generating empty skeletons for missing types. Following the Diataxis "work by improvement" principle, recommend the single highest-value document to add or fix next, and offer to draft it via draft-docs.

Notes

  • Use --report-only to skip generation.
  • Avoid test files unless they are test helpers.
  • Keep report output aligned with the language-specific standards in the reference file.
  • The Diataxis balance check is about the doc set as a whole; per-language symbol coverage and type balance are independent — report both.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Ensure Docs AI skill do?

Verify documentation coverage and generate missing docs interactively

Why use Ensure Docs on TypingMind?

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

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/existential-birds/beagle/tree/main/plugins/beagle-docs/skills/ensure-docs. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Ensure 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 Ensure Docs?

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

Is the Ensure Docs AI skill free?

Yes. It is published on GitHub by existential-birds under the Apache-2.0 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 👇