The Agent Skills Directory

[SAFE]: The skill's instructions provide a heuristic for ranking metaphors and analogies. No malicious patterns, obfuscation, or dangerous command executions were detected.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits a potential surface for indirect prompt injection as it ingests external data. 1. Ingestion points: The skill reads voice profiles from shared-context/voice-profile.md and shared-context/voices/{section}.md. 2. Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the ranking instructions. 3. Capability inventory: No dangerous capabilities (such as shell execution, file writing, or network access) are available to the skill. 4. Sanitization: No sanitization is performed on the ingested content. Despite the lack of markers, the absence of exploitable tools makes this surface safe.

voice-fitness-check