codebase-onboarding

Analyzes an unfamiliar codebase and generates an onboarding guide and starter CLAUDE.md.

Updated Mar 25, 2026
One-click install
npx skills add https://github.com/Femad-6/my-skills --skill codebase-onboarding-femad-6
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: codebase-onboarding
Source: https://github.com/Femad-6/my-skills/tree/main/.github/skills/codebase-onboarding
Command: npx skills add https://github.com/Femad-6/my-skills --skill codebase-onboarding-femad-6

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Joining a new project or opening an unfamiliar repository means hours of digging through files to understand the stack, architecture, and conventions. This Skill automates that reconnaissance and produces a structured onboarding guide plus a project-specific CLAUDE.md. ## Core Features & Use Cases - Four-Phase Analysis: Runs reconnaissance (manifests, frameworks, entry points, configs), architecture mapping, convention detection, and artifact generation. - Onboarding Guide Generation: Produces a scannable guide covering tech stack, directory map, request lifecycle, conventions, and common commands. - CLAUDE.md Generation: Creates or enhances a project-specific CLAUDE.md while preserving existing instructions and flagging what changed. - Use Case: A developer joins a team maintaining a Next.js app. They ask to be onboarded, and the Skill detects the TypeScript/Prisma/Jest stack, maps key directories, traces the request lifecycle, and writes a tailored CLAUDE.md to the repo root. ## Quick Start Ask the assistant to onboard you to this codebase and generate a CLAUDE.md for the project.

Frequently Asked Questions about codebase-onboarding

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a CLAUDE.md for an existing project?

Ask the assistant to generate a CLAUDE.md for the project. The Skill runs reconnaissance on manifests, configs, and directory structure, detects conventions, and writes a project-specific CLAUDE.md covering stack, commands, and code style.

How do I understand an unfamiliar codebase quickly?

Request onboarding for the repository. The Skill fingerprints the framework, maps key directories, traces one request from entry to response, and produces a scannable onboarding guide with entry points, conventions, and common commands.

Does it work with monorepos and different languages?

Yes. Reconnaissance checks for package.json, go.mod, Cargo.toml, pyproject.toml, pom.xml, Gemfile, and other manifests, then fingerprints frameworks like Next.js, Django, Flask, FastAPI, and Rails to identify the architecture pattern.

What happens if a CLAUDE.md already exists in the repo?

The existing CLAUDE.md is read first and enhanced rather than replaced. Project-specific instructions are preserved, and additions or changes are clearly called out so nothing is silently overwritten.

What are the limitations of automated codebase analysis?

Conventions that cannot be confidently detected are flagged as unknown rather than guessed. Repos with no commits or shallow git history skip git convention detection, and the guide intentionally omits exhaustive dependency listings.