Generate Comprehensive Codebase Onboarding Guides
Generates two onboarding docs from a codebase: a principal-level design-rationale guide and a step-by-step contributor guide.
Why it matters
Generate two distinct onboarding guides for any codebase: a principal-level deep dive into architecture and decisions, and a zero-to-hero guide for new contributors with practical setup and workflow instructions.
Outcomes
What it gets done
Detect primary programming language from build files.
Generate principal-level guide with architecture, design rationale, and decision logs.
Generate contributor guide with setup, workflow, and debugging steps.
Cite all claims with file paths and line numbers.
Install
Add it to your toolbox
Run in your project directory:
curl -fsSL https://spark.entire.vc/get/ag-wiki-onboarding | bash Overview
Wiki Onboarding Guide Generator
Generates two onboarding documents from a codebase via /deep-wiki:onboard: a Principal-Level guide covering architectural rationale and tradeoffs, and a Zero-to-Hero Contributor Guide with step-by-step setup and practical guidance. Use when asked for onboarding docs, a getting-started guide, or running /deep-wiki:onboard to help new team members understand a codebase.
What it does
This skill generates two complementary onboarding documents from a codebase, invoked via the /deep-wiki:onboard command, together covering engineers from newcomer to principal level. It first detects the repository's primary language from its build files (package.json/tsconfig.json for TypeScript/JavaScript, .csproj/.sln for C#/.NET, Cargo.toml for Rust, pyproject.toml/setup.py/requirements.txt for Python, go.mod for Go, pom.xml/build.gradle for Java) so all code examples match the actual stack.
The first document, a Principal-Level Onboarding guide for senior/staff+ engineers, requires twelve sections centered on the "why" behind decisions: system philosophy and invariants, an architecture overview with a component-map Mermaid diagram, the load-bearing key abstractions everything depends on, a decision log with alternatives considered and tradeoffs, dependency rationale (why each major dependency was chosen and what it replaced), data flow traced from actual code rather than guessed, failure modes and error propagation, performance characteristics and bottlenecks, the security model (auth, trust boundaries, data sensitivity), testing strategy including what's NOT tested, operational concerns (deployment, monitoring, feature flags), and an honest accounting of known technical debt. Every claim must carry a (file_path:line_number) citation, and the guide requires a minimum of three Mermaid diagrams (architecture, data flow, dependency graph) rendered in dark-mode colors.
The second document, a Zero-to-Hero Contributor Guide for new contributors, requires fourteen sections built around practical step-by-step guidance: a 2-3 sentence elevator pitch, prerequisites, environment setup with exact copy-pasteable commands and expected output at each step, an annotated project structure, an end-to-end first-task walkthrough, the development workflow (branching, commit conventions, PR process), how to run and add tests, a debugging guide for common issues, key domain concepts with code examples, reusable "if you want to add X, follow this pattern" code templates, common new-contributor pitfalls, where to get help, a glossary, and a quick-reference command cheat sheet - again grounded in (file_path:line_number) citations and dark-mode Mermaid diagrams for workflows.
When to use - and when NOT to
Use this skill when a user asks for onboarding docs or a getting-started guide, runs the /deep-wiki:onboard command, or wants to help new team members understand a codebase - producing both the deep architectural rationale a senior engineer needs and the concrete step-by-step path a newcomer needs, rather than a single document trying to serve both audiences.
It is a codebase-grounded documentation generator specifically - every claim is tied to a real file/line citation, so it's not suited to documenting a system that hasn't been implemented yet or producing purely aspirational architecture docs.
Inputs and outputs
Input is a repository to onboard someone onto. Output is two documents: a Principal-Level Onboarding guide (12 sections, 3+ Mermaid diagrams, decision-log style) and a Zero-to-Hero Contributor Guide (14 sections, copy-pasteable commands, glossary, cheat sheet), both citing (file_path:line_number) for every claim and using the detected primary language for code examples.
Integrations
Generates Mermaid diagrams using dark-mode colors per a companion wiki-vitepress skill's convention, and detects language/framework from standard build manifest files across TypeScript/JavaScript, C#/.NET, Rust, Python, Go, and Java ecosystems.
Who it's for
Teams onboarding both senior engineers, who need architectural rationale and tradeoffs, and new contributors, who need a concrete step-by-step path, onto an existing codebase, wanting both documents grounded in the actual code rather than written from memory or assumption.
FAQ
Common questions
Discussion
Questions & comments · 0
Sign In Sign in to leave a comment.