The Agent Skills Directory

[SAFE]: No security issues detected. The skill follows best practices for structured reasoning and contains no executable code or network exfiltration.- [PROMPT_INJECTION]: The skill processes user-defined decision statements that are passed to sub-agents, creating a surface for indirect prompt injection. 1. Ingestion points: User input for the decision statement in SKILL.md. 2. Boundary markers: Absent in the instructions provided to sub-agents. 3. Capability inventory: Spawning of isolated sub-agents for stance analysis. 4. Sanitization: Detailed instructions to adhere to ethical guardrails and refuse advocacy if a proposal is fundamentally harmful. This represents a functional risk associated with the skill's primary purpose of analyzing untrusted inputs.- [EXTERNAL_DOWNLOADS]: The skill references configuration content from a sibling repository path ('../perspectives/'). These are local file dependencies and no remote downloads from untrusted sources were detected.

consensus