What problem does it solve?
AI agents often struggle to maintain consistent, high-quality contextual information across complex projects. This skill provides a structured methodology for building REFERENCE.md files, ensuring that all agents have access to clear, hierarchical, and actionable context, reducing ambiguity and improving decision-making, so you can rest assured your AI is always informed.
Core Features & Use Cases
- Structured Reference Creation: Guides the creation of
REFERENCE.md files with mandatory ABSTRACT (one-paragraph summary) and CONTEXT sections, ensuring consistency and discoverability.
- Hierarchical Context Leveraging: Automatically accesses and stacks parent
REFERENCE.md abstracts, allowing child references to build upon existing knowledge without repetition, saving time and ensuring coherence.
- Quality Assurance: Provides a checklist for
DISCOVERY.md, ABSTRACT, and CONTEXT quality, ensuring references are information-dense, self-contained, and actionable for any agent.
- Use Case: An AI agent needs to document a new microservice. It uses this skill to create a
REFERENCE.md for the service, leveraging existing project-level REFERENCE.md context, and ensuring the new reference's ABSTRACT is concise and informative for other agents, making the documentation process efficient and reliable.
Quick Start
Use yggdrasil-bootstrap skill. I am in a _bootstrap/execution/ work cell. Create a REFERENCE.md file for the parent directory that describes a new API endpoint for user authentication. Ensure the ABSTRACT is information-dense.