init-deep

Generate hierarchical AGENTS.md knowledge bases for software repositories.

5|Updated May 17, 2026
One-click install
npx skills add https://github.com/islee23520/lfg --skill init-deep-islee23520
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: init-deep
Source: https://github.com/islee23520/lfg/tree/main/skills/init-deep
Command: npx skills add https://github.com/islee23520/lfg --skill init-deep-islee23520

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill turns a codebase into a hierarchical AGENTS.md knowledge base so teams can capture repository-specific guidance at the right directory levels instead of relying on scattered tribal knowledge.

Core Features & Use Cases

  • Creates a root AGENTS.md plus selective subdirectory files based on measured project complexity.
  • Uses concurrent discovery, structure analysis, and symbol-aware inspection to locate meaningful documentation boundaries.
  • Fits repositories that need maintainable onboarding notes, local conventions, anti-pattern warnings, and targeted workflow guidance for specific modules.

Quick Start

Ask the assistant to generate or update hierarchical AGENTS.md files for the repository using this skill’s discovery, scoring, and validation workflow.

Frequently Asked Questions about init-deep

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

FAQPage Schema
How do I generate hierarchical AGENTS.md files for a software repository?

Hierarchical AGENTS.md generation uses concurrent discovery, LSP, and codegraph analysis to map project complexity. It creates a root AGENTS.md and selective subdirectory files to capture project conventions and workflows.

What is the best way to document repository-specific conventions and onboarding notes?

The best way to document repository conventions is by building a layered AGENTS.md knowledge base. This maps precise directory boundaries and complexity, replacing scattered tribal knowledge with maintainable onboarding guidance.

How does LSP and codegraph analysis help map project complexity for documentation?

LSP and codegraph analysis map project complexity by performing symbol-aware inspection to locate meaningful documentation boundaries. This concurrent discovery ensures generated files are placed accurately based on measured structural complexity.

Can I use this approach to document local conventions and anti-pattern warnings for specific modules?

Yes, you can document local conventions and anti-pattern warnings by generating selective subdirectory AGENTS.md files. The scoring-based placement workflow targets specific modules requiring dedicated workflow guidance.

Does generating AGENTS.md files require deduplicated validation of the output?

Generating AGENTS.md files requires deduplicated validation to ensure the output knowledge base is precise and non-redundant. This validation step follows the scoring-based placement and concurrent discovery phases.