Analyzes codebases to generate and maintain a comprehensive project memory system for consistent AI context.
This skill acts as a project architect and historian for Claude Code, enabling it to automatically map out directory structures, identify tech stacks, and document business logic. By maintaining a standardized set of steering documents and memory records, such as Architectural Decision Records (ADRs) and domain-specific workflows, it ensures that every interaction with the codebase is informed by the project's core principles. This prevents context drift, facilitates multi-agent collaboration, and provides a persistent knowledge base that evolves alongside your code.
Key Features
011 GitHub stars
02Automated architecture pattern detection and tech stack extraction from source code.
03Bilingual documentation generation (English and Japanese) for structure, tech, and product context.
04Persistent memory management for ADRs, development workflows, and domain knowledge.
05Drift detection to identify inconsistencies between project documentation and actual code implementation.
06Interactive multi-mode dialogue for bootstrapping, syncing, and reviewing project metadata.
Use Cases
01Maintaining a living history of architectural decisions (ADRs) and lessons learned during development.
02Onboarding Claude to a complex legacy repository by automatically generating an architectural map.
03Ensuring technical documentation stays synchronized with package updates and structural refactors.
