Can I use Onboard on projects that already have a CLAUDE.md?

Yes, it is designed to merge with existing CLAUDE.md files, preserving manual human-written sections while updating auto-detected technical details.

What is the Context Budget Advisory?

It is an audit that checks if your current configuration (MCP tools and documentation) is too heavy, helping you optimize Claude's performance and token usage.

Does it read every single file in my repository?

No, it performs a targeted scan of configuration files, entry points, and 3-5 representative source files to extract patterns without exhausting your context window.

What files does the Onboard skill create?

It creates a CLAUDE.md file in the root and a .rune/ directory containing conventions, architectural decisions, progress logs, and a human-readable developer guide.

How does this save time during AI sessions?

By providing a persistent source of truth for your project's architecture and rules, Claude doesn't need to be re-taught your tech stack or conventions at the start of every new chat.

Onboard

Name: Onboard
Author: Rune-kit

byRune-kit

•

Aprendizaje y Documentación

Auto-generates comprehensive project context and documentation to sync AI agents with any codebase instantly.

The Onboard skill automates the critical first step of AI-assisted coding by scanning your codebase to establish deep project context. It detects tech stacks, naming conventions, and architectural patterns to generate a standardized CLAUDE.md and a persistent .rune state directory. This process eliminates the need to manually re-explain project details in every new session, saving developers significant time and ensuring that Claude operates with full knowledge of project-specific standards and environmental requirements.

Características Principales

01Automated tech stack and dependency detection

02Creation of human-readable DEVELOPER-GUIDE.md files

03Generation of standardized CLAUDE.md configuration files

0458 GitHub stars

05AI context budget auditing to optimize performance

06Extraction of project-specific naming and coding conventions

Casos de Uso

01Initializing a legacy or undocumented project for AI-assisted development

02Reducing context-window overhead by auditing MCP tools and documentation size

03Synchronizing project conventions and state across distributed team members

Características Principales

01Automated tech stack and dependency detection

02Creation of human-readable DEVELOPER-GUIDE.md files

03Generation of standardized CLAUDE.md configuration files

0458 GitHub stars

05AI context budget auditing to optimize performance

06Extraction of project-specific naming and coding conventions

Casos de Uso

01Initializing a legacy or undocumented project for AI-assisted development

02Reducing context-window overhead by auditing MCP tools and documentation size

03Synchronizing project conventions and state across distributed team members