uizze-ui-research

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs using an "Authorization: Bearer " header (and thus encourages embedding an agent token in requests/commands), which requires the LLM to handle or output secret values verbatim and creates exfiltration risk.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.75). The required workflow explicitly instructs the agent to “find relevant real web or iOS UI references” from the public UIZZE catalog (a runtime web-content source), meaning the LLM will likely ingest outsider-authored free text extracted from those references via the UIZZE tools/MCP.