What are the five discovery layers mentioned?

The layers are: CLAUDE.md (Quick Reference), capabilities.json (Structure), Skills Index (Workflows), Weaviate (Semantic Search), and Source Code (Implementation Details).

What is the primary source of truth for CodeCompass capabilities?

The .ai/capabilities.json file serves as the definitive map, detailing all domains, modules, and entry points within the platform.

When should I use the discover-capabilities skill?

Apply this skill at the start of a session, when a user asks 'What can CodeCompass do?', or before implementing new features to check for existing logic.

How does this skill improve token efficiency?

Instead of performing broad grep/glob searches across the codebase, it targets structured discovery layers, significantly reducing exploratory token consumption.

CodeCompass Capability Discovery

Name: CodeCompass Capability Discovery
Author: PearlThoughts

byPearlThoughts

生産性とワークフロー

Explores and maps the capabilities, modules, and architectural domains of the CodeCompass platform using structured metadata.

概要

This meta-skill provides a structured framework for navigating the expansive CodeCompass ecosystem. By leveraging a multi-layered discovery approach—ranging from high-level summaries in CLAUDE.md to granular module maps in capabilities.json—it eliminates inefficient manual exploration. It is essential for initial session onboarding, planning complex feature implementations, and identifying pre-existing modules across five distinct domains, including code analysis, semantic search, and requirements extraction.

主な機能

Multi-layered discovery spanning documentation, JSON maps, and semantic search
0 GitHub stars
Structured mapping of 5 architectural domains and 21+ modules
Dependency and relationship tracing between internal modules
Automated parsing of capabilities.json for instant system awareness
Decision tree logic to determine the most efficient discovery path

ユースケース

Pre-implementation research to identify and reuse existing modules
Initial session onboarding to understand the available toolset and environment
Complex workflow planning by mapping dependencies and entry points

概要

主な機能

Multi-layered discovery spanning documentation, JSON maps, and semantic search
0 GitHub stars
Structured mapping of 5 architectural domains and 21+ modules
Dependency and relationship tracing between internal modules
Automated parsing of capabilities.json for instant system awareness
Decision tree logic to determine the most efficient discovery path

ユースケース

Pre-implementation research to identify and reuse existing modules
Initial session onboarding to understand the available toolset and environment
Complex workflow planning by mapping dependencies and entry points