soul-brews-studio/opensource-nat-brain-oracle · Archived

learn

Explore a codebase with 2 parallel Haiku agents and create concise documentation. Use when user says "learn [repo]", "explore codebase", "study this repo", or wants to understand a project.

First seen Mar 28, 2026

Installation

$ npx skills add soul-brews-studio/opensource-nat-brain-oracle --skill learn

Stronger alternatives

This repository is archived — consider an actively maintained alternative.

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

Also in this package

Other skills from soul-brews-studio/opensource-nat-brain-oracle.

npx skills add soul-brews-studio/opensource-nat-brain-oracle

Browse all from soul-brews-studio/opensource-nat-brain-oracle

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Repository health

Stars 4
Default branch main
Open issues 5
Status Archived

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 3,285 B
  • docs SUMMARY.md 202 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 1 installs

SKILL.md

/learn - Codebase Learning

Explore a codebase → create hub + 3 focused docs.

Usage

/learn [url]             # Clone + explore
/learn [repo-path]       # Explore local

Step 0: Clone if URL

date "+🕐 %H:%M (%A %d %B %Y)"
ghq get [url]
mkdir -p ψ/learn/[REPO_NAME]

Step 1: Launch 2 Agents (PARALLEL)

Agent 1: Structure Scout

Explore [REPO_PATH]. Return:

## What It Does
[2-3 paragraphs: project purpose, why it exists, who it's for]

## How It Works
[Core mechanism in plain English. What's the key idea?]

## Architecture
| Component | Purpose |
|-----------|---------|
[7-10 rows]

[Explain relationships between components - 2-3 sentences]

## Dependencies
| Package | Why |
|---------|-----|
[Key deps with explanation]

STYLE: Tables for structure, prose for explanation. ~80 lines.

Agent 2: Code Hunter

Explore [REPO_PATH]. Return:

## Key Patterns
For each pattern (3 patterns):

### [Pattern Name]
**What**: [1-2 sentences]
**Why it's clever**: [1-2 sentences]
[Code block]

## Quick Reference
- **Install**: [command]
- **Key files**: [list]
- **Entry point**: [file]

STYLE: Explain the "why", not just show. ~70 lines.

Step 2: Main Agent Creates 4 Files

File 1: [REPO].md (Hub)

# [REPO] - Learning Notes

**Date**: [TODAY]
**Repo**: [URL]

## Summary
[1 paragraph from Agent 1]

## Key Insight
[The ONE pattern worth stealing from Agent 2]

## Files
- [[ARCHITECTURE]] - Structure + how it works
- [[CODE-SNIPPETS]] - Patterns worth stealing
- [[QUICK-REFERENCE]] - Quick lookup

## Links
- Repo: [url]
- Demo: [if any]

File 2: ARCHITECTURE.md

# [REPO] - Architecture

## What It Does
[From Agent 1 - 2-3 paragraphs]

## How It Works
[From Agent 1 - core mechanism]

## Components
| Component | Purpose |
|-----------|---------|
[From Agent 1]

[Relationship explanation]

## Dependencies
| Package | Why |
|---------|-----|
[From Agent 1]

File 3: CODE-SNIPPETS.md

# [REPO] - Code Snippets

## Key Patterns

### 1. [Pattern Name]
**What**: [explanation]
**Why**: [why it's clever]
[code block]

### 2. [Pattern Name]
...

### 3. [Pattern Name]
...

File 4: QUICK-REFERENCE.md

# [REPO] - Quick Reference

## Install
[command]

## Key Files
| File | Purpose |
|------|---------|
[5-7 rows]

## Entry Points
- Main: [file]
- Config: [file]

## Links
- Repo: [url]
- Docs: [if any]

Output Summary

## 📚 /learn: [REPO_NAME]

**What**: [1 sentence]
**Key Pattern**: [1 sentence]

**Files**:
- `[REPO].md` - Hub
- `ARCHITECTURE.md` - Structure
- `CODE-SNIPPETS.md` - Patterns
- `QUICK-REFERENCE.md` - Quick lookup

**Location**: ψ/learn/[REPO_NAME]/

Style Rules

  • Hub: ~40 lines (summary + links)
  • Architecture: ~80 lines (tables + explanation)
  • Code Snippets: ~70 lines (patterns + why)
  • Quick Reference: ~40 lines (lookup)
  • Total: ~230 lines across 4 files
  • Tables for structure, prose for explanation
  • English for technical content