council
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references a markdown file hosted on the author's own GitHub repository (csark0812) to provide shared composition vocabulary. This is a vendor-owned resource used for documentation purposes.- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources and task artifacts. It mitigates the risk of indirect prompt injection by including a non-negotiable instruction to treat all such data as untrusted evidence rather than instructions, and explicitly states that these inputs cannot authorize privileged operations such as tool use, code edits, or secret access.
Audit Metadata