One-click install
npx skills add https://github.com/INNERJOINT/HarnessSkills --skill deepinit-innerjoint
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deepinit
Source: https://github.com/INNERJOINT/HarnessSkills/tree/main/skills/deepinit
Command: npx skills add https://github.com/INNERJOINT/HarnessSkills --skill deepinit-innerjoint

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Large codebases lack concise, agent-friendly documentation that explains directory purpose, key files, relationships, and special instructions for automated agents, making onboarding, automated analysis, and multi-agent workflows slow and error-prone.

Core Features & Use Cases

  • Hierarchical AGENTS.md generation: produce AGENTS.md per directory with parent reference tags to create a navigable documentation tree.
  • Update mode and preservation: detect existing AGENTS.md, merge auto-generated updates while preserving manual notes and timestamps.
  • Validation and workflows: verify parent links, detect or delete orphaned AGENTS.md, and apply parallelization rules to balance performance.
  • Use cases include repository-wide onboarding docs, automated agent guidance for code modifications, and continuous documentation maintenance during CI runs.

Quick Start

Run the deepinit skill to scan the repository, map directory hierarchies, and generate or update AGENTS.md files with correct parent references and timestamps.

Frequently Asked Questions about deepinit

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

FAQPage Schema
How do I generate hierarchical documentation for a large codebase?

To generate hierarchical codebase documentation, you can scan your repository to map directory hierarchies and automatically create an AGENTS.md file for each directory with parent reference tags. This builds a navigable documentation tree for automated agents.

What is the best way to create AI-readable docs for automated agents?

AI-readable docs are structured files that explain directory purpose, key files, and relationships to guide automated agents. Generating AGENTS.md files across your repository provides concise, agent-friendly context for onboarding and multi-agent workflows.

Can I update existing AGENTS.md files without losing manual notes?

Yes, updating existing AGENTS.md files detects manual notes and preserves them during the merge. The generation process applies merge logic to auto-update content while keeping your manual sections and timestamps intact.

How do you validate parent links and handle orphaned files in generated docs?

Validating parent links involves verifying reference tags across the directory hierarchy and checking for orphaned AGENTS.md files. The process detects orphaned files and can delete them while applying parallelization rules to balance performance.

Does generating repository-wide documentation work for continuous integration runs?

Yes, generating repository-wide documentation works for continuous integration runs by applying configurable parallelization and empty-directory handling. This enables continuous documentation maintenance during CI workflows without manual intervention.