agentleFS
bbvch-aibbvch-ai/aihub-core/.claude/skills/document-feature/SKILL.md

document-feature

Skill16 starsChanged 3 days agoApache-2.0View on GitHub

Create user-facing feature documentation for the docs VitePress docs site. Analyzes the codebase and produces an index.en.md in the correct docs directory. Use when user says 'document this feature', 'write feature docs', 'create docs for X', 'add feature to docs site', or 'user-facing documentation for'. Takes feature name as argument. Do NOT use for ADRs (use /document-decision) or syncing existing docs (use /update-doc).

We split this skill into the rules it gives an agent

6 rules. 0 are common, 3 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Create user-facing documentation for feature "\$ARGUMENTS" on the Swiss AI Hub VitePress docs site.

    line 9
  2. Distinctive:

    Read these existing feature docs to understand the tone and structure:

    line 14
  3. Very specific:

    Search for the feature across scopes (packages/core, packages/agent, packages/api, packages/pipeline, packages/process, packages/web).

    line 26
  4. Very specific:

    Create index.en.md (NOT index.md) in the new directory.

    line 46
  5. Distinctive:

    No code blocks that will drift out of sync — reference file paths instead

    line 60
  6. Distinctive:

    After creating the English file:

    line 65

“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Document Feature - Create User-Facing VitePress Documentation

Create user-facing documentation for feature "$ARGUMENTS" on the Swiss AI Hub VitePress docs site. Analyzes the codebase, then produces an index.en.md file for end users.

Before You Start

Read these existing feature docs to understand the tone and structure:

  • MCP: docs/docs/2_platform/19_mcp/index.en.md — standards-based integration doc with capabilities breakdown
  • Memory: docs/docs/2_platform/15_memory/index.en.md — conversational style with "why it matters", "how it works", "two types" sections
  • Cost Control: docs/docs/2_platform/14_cost_control/index.en.md — uses ::: details containers organically, includes comparison tables

Real docs use natural, conversational section headers — NOT a rigid template. Match the style you see.

Step 1: Research the Feature

Search for the feature across scopes (packages/core, packages/agent, packages/api, packages/pipeline, packages/process, packages/web). Understand what it does, how users interact with it, and what makes it valuable.

Step 2: Choose the Docs Location

Feature docs live under docs/docs/. Determine where based on audience:

  • Platform features (user-facing: UI, chat, admin): docs/docs/2_platform/{N}_{topic}/index.en.md
  • SDK features (developer-facing: building agents, pipelines, processes): docs/docs/3_sdk/6_feature_overview/

Directory naming uses a numeric prefix for sidebar ordering (e.g., 19_mcp/, 15_memory/). Check existing directories to pick the next available number:

ls -d docs/docs/2_platform/*/
ls -d docs/docs/3_sdk/*/

Step 3: Create the Documentation File

Create index.en.md (NOT index.md) in the new directory.

Front matter — only title is required. Ordering comes from the directory prefix, not frontmatter:

---
title: Feature Title
---

Writing conventions (from docs/CLAUDE.md):

  • User-facing perspective, present tense, jargon-free language
  • VitePress containers: ::: info, ::: warning, ::: details — use organically, not as rigid template
  • No code blocks that will drift out of sync — reference file paths instead
  • Focus on what the feature does for users, not internal architecture

Step 4: Handle Translation

After creating the English file:

  • Never create index.de.md manually — it is auto-generated by pnpm run docs:translate
  • The translation pipeline uses SHA-256 hashes to detect changes (source_sha in .de.md frontmatter)
  • German translations must be committed to git (CI doesn't run translation)

Step 5: Verify

# File exists at the right location
ls docs/docs/2_platform/*_{topic}/index.en.md

# Front matter has title
head -5 docs/docs/2_platform/*_{topic}/index.en.md

# Preview (if dev server is available)
cd docs && pnpm run docs:dev

Compare your doc with the exemplars from "Before You Start" — it should match their tone and depth.

Troubleshooting

ProblemSolution
Unsure if platform or SDK featureUI/chat interaction = platform (2_platform/); code/API = SDK (3_sdk/)
Feature spans many scopesFocus on user-visible behavior, not internal architecture
Feature not yet fully implementedUse ::: warning boxes to note incomplete sections
Don't create docs in 6_code_deep_dive/That directory is auto-synced from READMEs via sync-docs.sh

Done When

  • index.en.md created in the correct docs/docs/ subdirectory
  • Front matter has title field
  • Style matches existing feature docs (MCP, Memory, Cost Control)
  • No index.de.md created manually (translation pipeline handles this)

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.