fengge-depressed
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes natural language input from users to generate responses in character, creating a vulnerability surface for indirect prompt injection.
- Ingestion points: User-provided text triggered by phrases like 'emo' or 'broken' as listed in SKILL.md.
- Boundary markers: The agent is not instructed to use delimiters or ignore instructions embedded within the user input.
- Capability inventory: No high-risk tools (file write, command execution, network access) are requested or utilized in the skill definition.
- Sanitization: Input validation or filtering is not implemented in the persona instructions.
- [SAFE]: The skill's external references point to reputable news organizations and information services for character research purposes. These references are static and do not trigger automated code execution.
Audit Metadata