learn

Orchestrate three Haiku agents to inspect codebases and generate organized documentation.

2|3|Updated Dec 21, 2025
One-click install
npx skills add https://github.com/laris-co/nat-agents-core --skill learn-laris-co
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: learn
Source: https://github.com/laris-co/nat-agents-core/tree/main/bundles/skills/learn
Command: npx skills add https://github.com/laris-co/nat-agents-core --skill learn-laris-co

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Automates the process of understanding a codebase by coordinating three Haiku agents to generate architecture summaries, code snippets, and a quick reference.

Core Features & Use Cases

  • Three parallel Haiku agents coordinate to map architecture, collect code snippets, and build a quick reference.
  • Generates a searchable hub of documentation under ψ/learn/[repo_name].
  • Use cases include rapid onboarding, codebase exploration, and project knowledge capture.

Quick Start

Start by running /learn with a repo URL or slug to begin automatic exploration and documentation.

Frequently Asked Questions about learn

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

FAQPage Schema
How do I generate documentation for a codebase from a GitHub repo?

Generate codebase documentation by providing a GitHub URL, local path, or slug. Three Haiku agents coordinate to map architecture, collect code snippets, and build a quick reference written to a dedicated folder.

Can I explore a local repository and auto-generate an architecture summary?

Explore local repositories by providing the local path. The agents inspect the codebase to auto-generate architecture summaries and quick references, outputting the organized documentation under a dedicated directory.

What is the best way to onboard new developers to an unfamiliar codebase?

Rapid onboarding is achieved by running a three-agent exploration on the repo. This coordinates agents to produce architecture summaries and code snippets, capturing project knowledge for new developers.

Does codebase exploration work with both URL and slug inputs?

Codebase exploration supports URL, slug, and local path inputs. It clones or accesses the target repo and orchestrates three parallel agents to document the architecture and code structure.

How does multi-agent codebase documentation write results to the workspace?

Multi-agent documentation writes results to a structured directory. The generated architecture summaries, code snippets, and quick references are saved as a searchable hub under a project-specific folder.

Do I need specific dependencies to automate codebase exploration?

No specific dependencies are required to automate codebase exploration. The process handles local repos and GitHub clones internally using three coordinated agents to generate the documentation.