omni-inference

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

• Third-party content exposure detected (low risk: 0.10). The workflow involves web fetching and search APIs which allow the agent to read arbitrary web pages and search results on demand.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The skill references an external GitHub raw URL for setup instructions without established first-party provenance, constituting a weak runtime dependency.