octocode-documentaion-writer_skill

Overview

This skill generates comprehensive repository documentation using a production-ready 6-phase pipeline with intelligent orchestration and conflict-free file ownership. It orchestrates specialized AI agents in parallel where appropriate, produces evidence-backed research, and outputs a validated documentation set and QA summary. It is built for TypeScript projects but supports multi-language codebases for analysis and docs generation.

How this skill works

The pipeline runs six phases: Discovery+Analysis, Question Engineering, Evidence Research, Orchestration and Assignment, Parallel Documentation Writers, and QA Validation. Parallel phases (1, 3, 5) spawn multiple agent tasks in a single message to ensure concurrent execution; outputs are aggregated into analysis.json, research.json, work-assignments.json and final documentation/*.md. Pre-flight checks and workspace initialization prevent running on build directories or invalid paths.

When to use it

• You need comprehensive, repo-level documentation generated automatically
• You want evidence-backed answers and citations from the codebase
• You need parallelized work to complete docs faster on medium-to-large projects
• You want conflict-free file ownership for multi-writer documentation
• You want an audited QA pass with LSP-powered verification

Best practices

• Run pre-flight checks and confirm repository path before spawning agents
• Always launch parallel agents for phases 1, 3, and 5 in a single message with all Task calls included
• Ensure exclusive file scopes for each writer to avoid conflicts
• If repository is very large (>200k LOC) acknowledge the warning and consider narrowing scope
• Resume from state.json only when you understand previous phase progress

Example use cases

• Generate full developer docs for a TypeScript monorepo with multiple services
• Create API and architecture documentation for a newly inherited codebase
• Produce evidence-backed component and flow documentation for onboarding new engineers
• Regenerate docs after major refactor using file-assigned parallel writers
• Perform QA validation against documentation to ensure code references are accurate