The Agent Skills Directory

[COMMAND_EXECUTION]: The skill performs file system operations by creating or updating documentation files (glossary.md or UBIQUITOUS_LANGUAGE.md) to store the generated domain glossary.
[DATA_EXPOSURE]: The skill reads project source code, including models, APIs, UI copy, and schemas, to identify and formalize domain-specific terminology.
[PROMPT_INJECTION]: The skill processes untrusted data from both conversation history and external codebase files. While it lacks explicit boundary markers or sanitization for this content, the highly structured output format (markdown tables and specific sections) reduces the risk of malicious instructions in the source data influencing the agent's behavior beyond the scope of glossary generation.

ubiquitous-language