# Docs Manager > Documentation management agent for the Pulse JS framework. Use this skill to collect markdown files from the docs directory, update documentation pages in docs/src, and maintain documentation consistency. Handles content migration, page updates, and cleanup of processed markdown files. Use when this capability is needed. - Skill: `tomevault-io/docs-manager-4` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/docs-manager-4` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/docs-manager-4/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/docs-manager-4 --- # Documentation Manager for Pulse Framework ## When to Use This Skill - Migrating content from .md files to documentation pages - Updating documentation in docs/src/pages/ from source .md files - Consolidating documentation updates across multiple files - Cleaning up processed markdown files after integration - Synchronizing documentation content with the docs website - Batch updating documentation pages from markdown sources ## Context Loading Load `.claude/context/export-map.md` for correct import paths. Load relevant API context file for the module being documented (see CLAUDE.md context table). ## Core Responsibilities ### 1. Documentation Collection - **Find markdown files**: Scan `docs/` directory for `.md` files (excluding `docs/src/`) - **Analyze content**: Read and understand the structure and purpose of each markdown file - **Map to pages**: Determine which documentation page(s) each markdown corresponds to ### 2. Page Updates - **Update pages**: Intelligently merge content into `docs/src/pages/*.js` files - **Preserve structure**: Maintain the existing page component structure and imports - **Add new content**: Integrate new sections, examples, and explanations - **Improve existing**: Enhance clarity, add missing details, update outdated information ### 3. Cleanup - **Delete processed files**: Remove `.md` files after successful integration - **Verify changes**: Ensure updates don't break the documentation site - **Track changes**: Log what was updated for review ## Documentation Structure ### Source Markdown Files (docs/) ``` docs/ ├── http.md → HttpPage.js ├── websocket.md → WebSocketPage.js ├── graphql.md → GraphQLPage.js ├── accessibility.md → AccessibilityPage.js ├── context.md → ContextPage.js ├── devtools.md → DevToolsPage.js ├── api.md → ApiReferencePage.js ├── cli.md → GettingStartedPage.js (CLI section) ├── pulse-dsl.md → CoreConceptsPage.js (.pulse DSL section) ├── internals.md → InternalsPage.js └── adr/ → Architecture Decision Records (keep) └── *.md → Referenced from InternalsPage.js ``` ### Documentation Pages (docs/src/pages/) ```javascript // Each page exports a component function: export function HttpPage() { return el('article.docs-page', [ el('h1', 'HTTP Client'), el('section.intro', [/* intro content */]), el('section.api', [/* API reference */]), el('section.examples', [/* code examples */]) ]); } ``` ## Update Strategy ### Step 1: Discovery ```bash # Find all markdown files (excluding ADRs and docs/src/) find docs -name "*.md" -not -path "docs/src/*" -not -path "docs/adr/*" ``` ### Step 2: Content Analysis For each markdown file: 1. **Read content**: Parse headings, sections, code blocks 2. **Identify target page**: Match filename to page (e.g., `http.md` → `HttpPage.js`) 3. **Extract sections**: Categorize content (intro, API, examples, etc.) ### Step 3: Page Update For each target page: 1. **Read existing page**: Understand current structure 2. **Merge content**: - Add new sections as `el('section', ...)` - Update existing text with improvements - Add code examples as `el('pre', el('code', ...))` - Preserve existing examples and structure 3. **Maintain imports**: Keep all necessary imports at the top 4. **Use syntax highlighting**: Wrap code in `highlightCode(code, 'javascript')` ### Step 4: Verification Before deleting source files: 1. **Check syntax**: Ensure the updated page is valid JavaScript 2. **Preview locally**: Run `pulse dev` to verify the page renders correctly 3. **Compare content**: Ensure all important content from .md is in the page ### Step 5: Cleanup After successful integration: 1. **Delete markdown files**: Remove processed `.md` files 2. **Keep ADRs**: Never delete `docs/adr/*.md` (architectural decisions) 3. **Keep milestones**: Preserve `docs/milestones/*.md` if they exist 4. **Log changes**: Report which files were processed ## Content Mapping Guidelines ### Common Sections to Map | Markdown Section | Page Component | |------------------|----------------| | `# Title` | `el('h1', 'Title')` | | `## Heading` | `el('h2', 'Heading')` | | Paragraph text | `el('p', '...')` | | Code block | `el('pre', el('code.language-js', highlightCode(...)))` | | Bullet list | `el('ul', items.map(item => el('li', item)))` | | Table | Convert to `el('table', ...)` with rows | | Note/Warning | `el('.note', ...)` or `el('.warning', ...)` | ### Code Example Pattern ```javascript import { highlightCode } from '../highlighter.js'; el('section.examples', [ el('h2', 'Examples'), el('h3', 'Basic Usage'), el('pre', el('code.language-javascript', highlightCode(` import { createHttp } from 'pulse-js-framework/runtime/http'; const api = createHttp({ baseURL: '/api' }); const data = await api.get('/users'); `, 'javascript') )) ]); ``` ### Internationalization (i18n) When updating pages, respect the i18n structure: ```javascript import { t } from '../i18n/index.js'; el('h1', () => t('pages.http.title')), // Reactive translation el('p', () => t('pages.http.intro')) ``` ## Files to Never Delete | Pattern | Reason | |---------|--------| | `docs/adr/*.md` | Architecture Decision Records (permanent reference) | | `docs/milestones/*.md` | Project planning documents | | `docs/src/**/*` | Documentation website source code | | `CLAUDE.md` | Project instructions | | `README.md` | GitHub repository readme | ## Workflow Example ### Input ``` docs/http.md exists with: - Introduction to HTTP client - API reference for createHttp() - Examples for GET/POST/interceptors ``` ### Process 1. **Read** `docs/http.md` 2. **Identify** target: `docs/src/pages/HttpPage.js` 3. **Read** `HttpPage.js` to understand current structure 4. **Update** `HttpPage.js`: - Add missing examples from markdown - Enhance API descriptions - Preserve existing structure 5. **Verify** page renders correctly 6. **Delete** `docs/http.md` ### Output ``` ✓ Updated HttpPage.js with content from http.md ✓ Added 3 new examples ✓ Enhanced API reference section ✓ Deleted docs/http.md ``` ## Error Handling ### If page doesn't exist - Create new page using template from `assets/page-template.js` - Add page import to `docs/src/state.js` - Add route to navigation ### If content is unclear - Ask for clarification about which page to update - Suggest multiple options if multiple pages could apply - Preserve original .md file until confirmed ### If page syntax breaks - Revert changes - Report error - Keep .md file for manual review ## Best Practices 1. **Preserve user content**: Never delete information that exists in pages 2. **Enhance, don't replace**: Add to existing content, improve clarity 3. **Maintain consistency**: Follow existing code style in pages 4. **Test before delete**: Always verify changes work before removing source files 5. **Batch operations**: Process all .md files in one session for consistency 6. **Log thoroughly**: Report all changes for user review ## Commands This skill responds to: - "Update documentation from markdown files" - "Migrate docs/*.md to docs/src/pages/" - "Sync documentation" - "Process pending documentation updates" After completion, the skill will: - List all updated pages - List all deleted markdown files - Provide a summary of changes - Suggest running `pulse dev` to preview changes --- > Converted and distributed by [TomeVault](https://tomevault.io/claim/vincenthirtz) — claim your Tome and manage your conversions.