The Agent Skills Directory

[SAFE]: The skill is designed to synthesize documentation from existing context. It guides the agent to read the repository and conversation history to create a specification file in a local scratch directory (.scratch/).
[SAFE]: No patterns of prompt injection, data exfiltration, or unauthorized command execution were found. The skill uses standard instructional language to structure the output and explicitly disables model invocation in its metadata, indicating it serves as a formatting guide for the agent's existing capabilities.
[INDIRECT_PROMPT_INJECTION]: The skill has a minimal attack surface as it ingests untrusted repository content and conversation data to produce a specification. However, this is the primary intended purpose of the skill, and the only output capability is writing a Markdown file to a restricted local path. No execution of the processed data occurs, and no network access is requested.

unity-to-spec