icon-retrieval
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill performs network requests using curl to the external domain lab.weavefox.cn, which is not included in the whitelisted or well-known service lists.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from an external search API and subsequent URL retrievals, representing a vulnerability surface. 1. Ingestion points: The skill reads data from the search API response at lab.weavefox.cn and the SVG content from resulting URLs. 2. Boundary markers: There are no instructions or delimiters defining boundaries for external data to prevent the agent from following embedded instructions. 3. Capability inventory: The agent has the capability to execute curl commands to fetch content from arbitrary URLs provided by the external API. 4. Sanitization: No sanitization, validation, or filtering of the external API responses or retrieved SVG content is specified.
Audit Metadata