The Agent Skills Directory

[SAFE]: The skill consists of instructions and markdown templates for domain modeling, glossary maintenance, and Architectural Decision Records (ADRs). No malicious patterns such as prompt injection, obfuscation, or remote code execution were detected.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it is instructed to cross-reference domain terms with the project's source code. This involves reading untrusted content from the codebase to validate definitions. However, the associated risk is negligible as the skill's capabilities are limited to providing feedback to the user and updating markdown documentation files.

domain-modeling