The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or dangerous commands were detected. The skill uses standard instructional language to define a specific agent persona for plan review.
[INDIRECT_PROMPT_INJECTION]: The skill includes instructions to ingest and process user plans and codebase contents (ingestion points). It does not explicitly define boundary markers or sanitization methods for this external data. However, the identified capabilities are limited to conversation and read-only codebase exploration, which align with the skill's intended purpose and present a low security risk profile.

grill-me