cold-shower
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted user-provided content such as technical plans, product ideas, and business narratives without specifying explicit boundary markers or delimiters to isolate data from instructions.
- Ingestion points: Processes user proposals, plans, and decisions provided in the chat context (SKILL.md).
- Boundary markers: None (the prompt instructions do not define delimiters for user data).
- Capability inventory: Limited to text analysis and the use of the environment's research/web tools for factual verification.
- Sanitization: No specific sanitization or filtering logic is mentioned for processed data.
- [EXTERNAL_DOWNLOADS]: The skill instructions direct the agent to use built-in environment research or web search tools when current facts about markets or competitors are required. This constitutes legitimate usage of the agent's capabilities for information gathering and does not involve downloading untrusted scripts or software.
Audit Metadata