wiki-architect_skill

Overview

This skill analyzes code repositories and generates hierarchical wiki catalogues and onboarding guides tailored to the project's structure and audience. It produces a JSON documentation scaffold mapping architecture layers, contributor paths, and role-specific guides to accelerate onboarding and maintain a single source of truth. Use it to create a discoverable docs site, a repo README expansion, or an internal wiki.

How this skill works

The skill first resolves the repository source context (detects remote URL and default branch) and then scans the file tree and README to detect languages, frameworks, and architecture patterns. It classifies layers (presentation, business logic, data, infra), derives titles from real files, and emits a constrained hierarchical JSON catalogue including onboarding, getting started, and deep-dive sections with file-level citations. Output respects nesting and size constraints and adapts for small repos.

When to use it

• You need a structured wiki or table-of-contents for a codebase
• Onboarding new contributors, staff engineers, product managers, or executives
• Producing a documentation plan or migration-ready docs scaffold
• Auditing codebase architecture or mapping system components
• Preparing role-specific guides (Contributor, Staff Engineer, Executive, PM)

Best practices

• Resolve the repository context (remote URL and branch) before analysis
• Keep the repository README and key files (main, build, config) present to improve title derivation
• Limit large deep-dives to clear component boundaries to respect nesting limits
• Cite real files in prompts so generated content links back to source code
• Generate separate role-specific appendices to avoid mixing audience-level details

Example use cases

• Generate a wiki scaffold for a Go microservice repository to speed new-hire ramp-up
• Create a role-based onboarding bundle (Contributor → Staff → Exec → PM) for an internal platform
• Map a mixed-language codebase into presentation, business, data, and infra layers for architecture reviews
• Produce a Getting Started and Deep Dive JSON catalogue to drive an automated docs site generator