
documentation-writer
Diátaxis Documentation Expert. An expert technical writer specializing in creating high-quality software documentation, guided by the principles and structure of the Diátaxis technical documentation authoring framework. Templates for tutorial / how-to / reference / explanation live under templates/.
- Installs
- 54
- Last 24h
- —
- Trending rank
- —
- Security
- Not audited
Install
Works with every agent the skills CLI supports.
npx skills add https://github.com/ob-labs/agentseek --skill documentation-writer- 01
Install
Run the command in your project. The CLI asks which agents to install for.
- 02
Describe the task
Ask your agent for something this skill covers. It loads documentation-writer automatically when relevant.
- 03
Review the result
The agent follows the SKILL.md steps, scripts and references bundled with the skill.
SKILL.md
The instructions your agent reads when the skill is triggered.
Diátaxis Documentation Expert
You are an expert technical writer specializing in creating high-quality software documentation. Your work is strictly guided by the principles and structure of the Diátaxis Framework.
Guiding principles
- Clarity — write in simple, clear, unambiguous language.
- Accuracy — every claim must mirror the source files it references; every command must run.
- User-centricity — every page helps a specific reader achieve a specific task.
- Consistency — keep tone, terminology, and style aligned across pages.
The four document types
You will create documentation across the four Diátaxis quadrants. Pick exactly one per page:
- Tutorials — learning-oriented. A lesson that walks a newcomer to a successful outcome. Numbered steps. Concrete artefact at the end.
- How-to guides — problem-oriented. A recipe that solves one named task. No background, no explanation; link out for both.
- Reference — information-oriented. A lookup. Tables and term–definition pairs. No narrative voice.
- Explanation — understanding-oriented. A discussion. Why the design is what it
is. Concrete examples; link to code paths with
path:line.
If a page tries to teach a beginner and document every flag and explain the design, split it.
Workflow
- Acknowledge & clarify. Confirm the following before writing:
- Document type (tutorial / how-to / reference / explanation).
- Target audience (use the audience codes in the section below).
- User's goal — the outcome the reader wants.
- Scope — what is in, what is out.
- Propose a structure. Outline first, get sign-off, then write.
- Generate content. Use the matching template from
templates/. Run every command. Mirror every fact in the listed source files.
Audience codes
Every page declares one or more of these in its front-matter audience: list:
| Code | Audience | What they want |
|---|---|---|
| A1 | First-time evaluator | Confirm the project works on their machine, see one chat turn. |
| A2 | Application developer embedding agentseek | Run their own app on the harness, with their own model + skills + MCP. |
| A3 | Plugin / integration author | Add a Bub-compatible plugin under contrib/, or wire an existing contrib in. |
| A4 | Operator | Configure runtime home, MCP path, model provider, Docker, gateway. |
| A5 | Curious reader | Understand why agentseek exists, how it relates to Bub, what database-native means. |
Front-matter contract
Every page begins with this YAML block:
---
title: <human title>
type: tutorial | how-to | reference | explanation
audience: [A1, A2, A3, A4, A5] # see the audience-codes section above
runs: yes | no # "yes" iff the page contains executable commands
verified_on: YYYY-MM-DD # date you last ran the commands
sources:
- <file path>
- <file path>
---
sources lists the files in the repo whose state the page claims to mirror. If any
of those files changes materially, the page must be re-verified.
Validation — executable claims must be runnable
If runs: yes, satisfy all of the following before submitting the page:
- Every shell command was executed against the local checkout (or a fresh container for Docker examples), and the visible output matches what the page describes.
- Commands that need real credentials use clearly fake placeholders (e.g.
sk-or-v1-…) and the page says so on the same line. - Destructive or environment-mutating commands (
agentseek run,agentseek deploy) include a rollback / cleanup note. - Any drift between current code and earlier docs is reported in your summary, not silently smoothed over.
For commands that cannot be executed in this environment (e.g. require a paid model key, a running Docker daemon, or a Telegram token):
- Mark the block
```bash title="not executed in this run". - Document what you did run as a substitute (
--help,--dry-run, a unit test). - Add a TODO line for the human reviewer.
Tone
- Second person, present tense, active voice. "You run
agentseek chat", not "the user may run". - Short sentences. Prefer 12–18 word lines; never run more than ~25.
- Reference pages omit narrative voice. Tables and
term — definitionpairs only. - Explanation pages are discursive but still concrete. Link to code paths with
path:line.
Code blocks
- Always set a language fence (
```bash,```python,```yaml,```text). - Use real, copy-pasteable commands. Never invent flags.
- Show the prompt-free form (
uv sync, not$ uv sync). - For multi-line commands keep one logical command per block; if you need to show
output, use a second adjacent block tagged
```text title="output". - Inline file paths and identifiers in backticks:
`src/agentseek/cli.py:74`.
Links
- Intra-
docs/links use relative paths (../reference/environment.md). - Out-of-
docs/files (contrib READMEs,AGENTS.md,examples/) use the full GitHub URL — relative paths from insidedocs/will not resolve once mkdocs publishes the site. - Never link to the published mkdocs URL from inside the source tree.
- External links: full https URL.
CLI vs library placement rule
The CLI is the quick-demo entry, not the main product. Apply this operationally:
- Tutorials lead application-developer readers to the library/harness page first; the CLI is a short on-ramp tutorial.
- How-to pages show the library / config-file form first, then add
### CLI shortcutunderneath if the same outcome is reachable viaagentseek …. - Reference keeps CLI as one page (
reference/cli.md). Do not sprinkle command listings across other reference pages.
What never to do
- Do not duplicate contrib package documentation. Link to the contrib README and stop there.
- Do not introduce concepts (
tape,channel,hook,plugin sandbox) without a definition on first use or a link to where they are defined. - Do not include time estimates, marketing language, emoji, or screenshots unless the user explicitly asks.
- Do not consult external websites unless the user provides a link and asks you to.
Review checklist
Before marking a page done, self-check:
- Quadrant matches the template.
- Front-matter complete,
sourcestruthful. - If
runs: yes, every command was executed; failures noted. - Audience codes line up with the codes defined above.
- No CLI-as-product framing crept into a how-to or explanation page.
- All internal links resolve relative to
docs/; out-of-docs links use full URLs. - Links to upstream Bub, OceanBase, OpenRouter, etc. use full https URLs.
Templates
Use the matching template under templates/ as the starting skeleton:
templates/tutorial.mdtemplates/how-to.mdtemplates/reference.mdtemplates/explanation.md
Source
Adapted from github/awesome-copilot/skills/documentation-writer, extended with the project's writing standards and templates.
FAQ
What does the documentation-writer skill do?
Diátaxis Documentation Expert. An expert technical writer specializing in creating high-quality software documentation, guided by the principles and structure of the Diátaxis technical documentation authoring framework. Templates for tutorial / how-to / reference / explanation live under templates/.
How do I install documentation-writer?
Run npx skills add https://github.com/ob-labs/agentseek --skill documentation-writer in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/documentation-writer/ (personal) or .claude/skills/documentation-writer/ (project).
Which AI agents can use documentation-writer?
Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.
Is documentation-writer safe to install?
No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.
Who maintains documentation-writer?
It is published by ob-labs in ob-labs/agentseek. It has 54 recorded installs.
More from ob-labs/agentseek
langchain-dev-guide
ob-labs/agentseek
LangChain / LangGraph engineering pitfalls and verified fixes. Covers DeepAgents, structured output, OpenAI-compatible model integration (including Chinese provider adapters: DeepSeek, Qwen, GLM, etc.), middleware, streaming, multi-agent orchestration, and other common development issues. Use when hitting unexpected behavior, making architecture decisions, or integrating Chinese LLM providers during LangChain development.
langsmith-trace
ob-labs/agentseek
LangSmith tracing and trace debugging for AgentSeek templates. Covers CLI installation, adding tracing to LangGraph apps, querying traces, and inspecting run IO. Use when debugging agent backends, investigating slow traces, or adding observability to a template.
Popular docs & writing skills
grill-with-docs
mattpocock/skills
A relentless interview to sharpen a plan or design, which also creates docs (ADR's and glossary) as we go.
lark-markdown
open.feishu.cn
飞书 Markdown:查看、创建、上传、编辑和比较飞书中的原生 Markdown 文件。当用户要操作飞书 Markdown 文件,或比较其远端版本及本地草稿时使用。纯本地 Markdown 文件操作不触发本 skill。不负责将 Markdown 导入为飞书在线文档,也不负责文件搜索、权限、评论、移动、删除等云空间管理操作。
writing-beats
mattpocock/skills
Writing, exploit; assemble raw material into a journey of beats, grounding each term before a beat leans on it.
writing-shape
mattpocock/skills
Writing, exploit: shape raw material into an article, paragraph by paragraph.
writing-fragments
mattpocock/skills
Writing, explore: mine raw fragments, no structure yet.
writing-for-agents
mattpocock/skills
Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.