add-flashcards
Pass
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: Indirect prompt injection surface identified.
- Ingestion points: The agent is instructed to fetch and process content from arbitrary user-provided URLs and workspace documents (SKILL.md).
- Boundary markers: The skill lacks explicit delimiters or instructions for the agent to ignore embedded commands within the source material.
- Capability inventory: The agent is granted permission to write files to the /home/workspace/flashcards/ directory (SKILL.md).
- Sanitization: No specific requirements for input validation or sanitization of the external source content are described.
- [PROMPT_INJECTION]: The skill instructions explicitly direct the agent to bypass user confirmation steps ("Zo should not ask for confirmation to continue"). This increases the agent's autonomy and reduces the user's ability to review potentially malicious file-writing operations triggered by external content.
Audit Metadata