How does the feature discovery phase work?

It scans your directory structure and config files to propose modules, then dispatches parallel subagents to identify specific user-facing capabilities and internal symbols.

Can I run the phases independently?

Yes. You can run Phase 1 to populate your feature list, skip to Phase 2 to map specific workflows manually, or run both for a comprehensive bootstrap.

How does it handle complex code paths?

It uses subagent tracing to follow code paths from entry points like API routes to final outcomes, then synthesizes these into clean Mermaid diagrams.

What is the difference between a Feature and Infrastructure?

Features are atomic, user-facing capabilities (like AUTH.login), while Infrastructure refers to shared internal utilities (like INFRA.logger) used by multiple features.

Feature Tree Bootstrap

Name: Feature Tree Bootstrap
Author: Nothflare

byNothflare

•

開発者ツール

Analyzes existing codebases to automatically discover atomic features and map complex user workflows.

概要

Feature Tree Bootstrap is a powerful diagnostic skill designed to reverse-engineer any codebase into a structured, AI-readable map. Using a two-phase process, it first performs a bottom-up analysis to identify user-facing features versus internal infrastructure, followed by a top-down trace that connects entry points to specific code symbols. This skill is essential for developers onboarding to legacy projects or those looking to create a definitive source of truth for AI-assisted development, ensuring every architectural layer is accounted for and documented.

主な機能

Top-down workflow tracing from routes and entry points
1 GitHub stars
Parallel subagent dispatch for rapid codebase scanning
Visual workflow synthesis using Mermaid diagram generation
Separation of user-facing features from shared infrastructure (INFRA)
Automated module detection and bottom-up feature discovery

ユースケース

Mapping code dependencies for AI-driven feature implementation
Onboarding onto a legacy codebase to understand architecture
Auditing architectural coverage to identify undocumented logic

概要

主な機能

Top-down workflow tracing from routes and entry points
1 GitHub stars
Parallel subagent dispatch for rapid codebase scanning
Visual workflow synthesis using Mermaid diagram generation
Separation of user-facing features from shared infrastructure (INFRA)
Automated module detection and bottom-up feature discovery

ユースケース

Mapping code dependencies for AI-driven feature implementation
Onboarding onto a legacy codebase to understand architecture
Auditing architectural coverage to identify undocumented logic