The Agent Skills Directory

[SAFE]: The skill's primary function is to parse conversation history and generate a structured Markdown file (UBIQUITOUS_LANGUAGE.md) containing tables and descriptions of domain terms.
[SAFE]: Analysis of the instructions shows no evidence of network exfiltration, sensitive file access (such as credentials or SSH keys), or persistence mechanisms.
[SAFE]: While the skill processes untrusted user input (conversation history), the output is constrained to a specific Markdown format with strict rules (e.g., 'Only include domain terms', 'Keep definitions tight'), which mitigates the risk of indirect prompt injection. The skill has no capability to execute the content it processes.
[SAFE]: The documentation includes an example dialogue mentioning shell commands and filesystem layers; however, these are used as conceptual domain examples for a hypothetical software system and do not represent executable instructions for the AI agent.
[SAFE]: The frontmatter configuration disable-model-invocation: true further restricts the agent's capabilities while the skill is active, reducing the overall attack surface.

ubiquitous-language