cocktail-mixology-master
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill establishes an automated research and update workflow in
modules/research-checklist.mdthat directs the agent to ingest data from external sources and use it to modify the skill's own guidance files, creating an indirect prompt injection surface. - Ingestion points: External data is fetched via web searches and distillation tools (Wikipedia, ArXiv) mentioned in the research checklist.
- Boundary markers: The instructions lack explicit boundary markers or warnings for the agent to disregard potential instructions embedded within the researched material.
- Capability inventory: The workflow requires the agent to update and overwrite its own skill files (such as
modules/core-guidance.md), necessitating file-write capabilities. - Sanitization: No validation or sanitization steps are defined to filter external content before it is incorporated into the skill's core instructions.
Audit Metadata