skill:docs-workflow - Documentation Lifecycle Manager

Version: 1.0.0

Purpose

The docs-workflow skill manages the complete documentation lifecycle for any project. It provides four slash commands — /docs , /docs-init , /docs-update , and /docs-claude — that create, maintain, and audit CLAUDE.md, README.md, and docs/ structures with smart templates. Like Hephaestus inscribing the purpose of each forged artifact, this skill ensures every project has clear, current, and comprehensive documentation.

File Structure

skills/docs-workflow/ ├── SKILL.md (this file) └── examples.md

Interface References

• Context: Loaded via ContextProvider Interface

• Memory: Accessed via MemoryStore Interface

• Schemas: Validated against context_metadata.schema.json and memory_entry.schema.json

Triggers

• create CLAUDE.md

• initialize documentation

• docs init

• update documentation

Mandatory Workflow

IMPORTANT: Execute ALL steps in order. Do not skip any step.

Step 1: Initial Analysis

• Detect documentation scope from user request and project state

• Identify project type (language, framework, monorepo vs single-repo)

• Determine which docs exist vs need creation:

• Check for CLAUDE.md , README.md , docs/ directory

• Identify the slash command being invoked (/docs , /docs-init , /docs-update , /docs-claude )

• Assess project structure to inform template selection

Step 2: Load Memory

• Use memoryStore.getSkillMemory("docs-workflow", "{project-name}") to load existing project memory. See MemoryStore Interface.

• Load project-specific documentation conventions and preferences

• Retrieve prior documentation generation history and patterns

• Check for established doc structure decisions from previous runs

Step 3: Load Context

• Load engineering domain context via contextProvider.getIndex("engineering") . See ContextProvider Interface.

• Load documentation standards and best practices

• Load project-type-specific templates and conventions

• Load output naming conventions from ../OUTPUT_CONVENTIONS.md

Step 4: Assess Documentation State

• Audit existing documentation files:

• CLAUDE.md: Check for project conventions, build/test commands, architecture notes, coding standards

• README.md: Check for project description, setup instructions, usage examples, contribution guidelines

• docs/ directory: Check for structure, completeness, and organization

• Identify gaps: missing sections, outdated content, incomplete instructions

• Compare current project state against documentation accuracy

• Generate a documentation health score

Step 5: Execute Documentation Action

Based on the command invoked, perform the appropriate action:

/docs-init — Initialize Documentation Structure

• Create CLAUDE.md with project conventions, build commands, and architecture overview

• Create or update README.md with project description, setup, usage, and contribution sections

• Create docs/ directory with appropriate structure for the project type

• Populate templates with detected project information (language, framework, dependencies)

/docs-update — Update Existing Documentation

• Scan project for changes since last documentation update (new files, changed APIs, updated dependencies)

• Update CLAUDE.md with current build/test commands and conventions

• Update README.md with current project state and any new sections needed

• Update docs/ files to reflect current architecture and APIs

• Flag sections that may need manual review

/docs-claude — Generate or Refresh CLAUDE.md

• Analyze project structure, build system, and conventions

• Generate or refresh CLAUDE.md with:

• Build and test commands

• Code style and conventions

• Architecture overview

• Key file locations

• Development workflow notes

• Preserve any manually-added sections from existing CLAUDE.md

/docs — Documentation Status and Health Summary

• Scan all documentation files for completeness and freshness

• Report documentation health: what exists, what's missing, what's outdated

• Provide actionable recommendations for improvement

• Display a summary dashboard with coverage metrics

Step 6: Generate Output

• Save output to /claudedocs/docs_workflow_{project}_{YYYY-MM-DD}.md

• Follow naming conventions in ../OUTPUT_CONVENTIONS.md

• Include summary of actions taken, files created/modified, and recommendations

Step 7: Update Memory

• Use memoryStore.update(layer="skill-specific", skill="docs-workflow", project="{project-name}", ...) to store insights. See MemoryStore Interface.

• Record documentation conventions discovered for this project

• Track which templates and structures were used

• Store documentation health history for trend tracking

• Note any manual overrides or custom preferences

Compliance Checklist

Before completing, verify:

• All mandatory workflow steps executed in order

• Memory loaded via memoryStore.getSkillMemory() (Step 2)

• Context loaded via contextProvider.getIndex() (Step 3)

• Documentation state assessed with gap analysis (Step 4)

• Correct command action executed (Step 5)

• Output saved with standard naming convention

• Memory updated via memoryStore.update() (Step 7)

Output File Naming Convention

Format: docs_workflow_{project}_{YYYY-MM-DD}.md

Where:

• {project} = Project name (lowercase, hyphens for spaces)

• {YYYY-MM-DD} = Date of execution

Examples:

• docs_workflow_my-api_2026-03-15.md

• docs_workflow_forge-plugin_2026-03-15.md

Version History

Version Date Changes

1.0.0 2026-02-12 Initial release — four-command documentation lifecycle with smart templates, gap analysis, and health scoring