The Agent Skills Directory

[SAFE]: The skill's primary purpose is domain modeling and documentation maintenance. It lacks functionality for code execution, network communication, or unauthorized system access.
[INDIRECT_PROMPT_INJECTION]: The skill processes project-specific files like GLOSSARY.md and source code, creating a surface for indirect prompt injection from local data.
Ingestion points: Reads content from GLOSSARY.md, GLOSSARY-MAP.md, and project files in the src/ directory.
Boundary markers: No explicit delimiters or instructions are used to distinguish between system prompts and document content.
Capability inventory: Operations are limited to reading and writing Markdown files. No shell or network tools are requested or used.
Sanitization: Content from files is processed without explicit sanitization.

domain-modeling