skills/freedomintelligence/openclaw-medical-skills/tooluniverse-immunotherapy-response-prediction/Gen Agent Trust Hub
tooluniverse-immunotherapy-response-prediction
Pass
Audited by Gen Agent Trust Hub on Jul 30, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted patient data (e.g., cancer types, mutation lists, and biomarker values) and interpolates it into complex analytical prompts, which creates a vulnerability surface for indirect prompt injection.\n
- Ingestion points: Patient tumor profiles provided as input and parsed according to the formats defined in
SKILL.md.\n - Boundary markers: None identified. The instructions do not define clear delimiters (e.g., XML tags or triple backticks) or provide explicit 'ignore' instructions to prevent the agent from following directives embedded in user-supplied medical data.\n
- Capability inventory: The skill utilizes a large array of informational tools for genomic, clinical, and pharmacological data retrieval (OpenTargets, MyGene, FDA, PubMed, etc.). It lacks active system-level capabilities such as shell access, arbitrary code execution, or persistent file system modification.\n
- Sanitization: No explicit validation or sanitization of input strings is performed before they are utilized as parameters for tool calls.
Audit Metadata