theme-alchemy
Pass
Audited by Gen Agent Trust Hub on Mar 14, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill uses a high-pressure roleplay persona ('PUA' style) in
SKILL.mdwhich employs aggressive imperatives to dictate agent behavior. This is a stylistic instruction override. - [PROMPT_INJECTION]: Vulnerability to indirect prompt injection exists due to the ingestion of untrusted data.
- Ingestion points: Variables
{{主药材}},{{辅药材}}, and{{丹药品级 / 杂质 < 1%}}inSKILL.md. - Boundary markers: Absent; user-controlled data is interpolated directly into the system prompt without delimiters or 'ignore' instructions.
- Capability inventory: None; the skill does not use tools, perform network operations, or execute code.
- Sanitization: Absent; there is no evidence of input validation or escaping for the ingested variables.
Audit Metadata