google-docs

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

• Third-party content exposure detected (low risk: 0.10). The agent reads user and shared Google Docs files via scripts/docs.py, which ingestion can include outsider-authored content depending on document sharing.

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

• Potentially malicious external URL detected (high risk: 0.70). The script fetches runtime OAuth refresh tokens from a third-party server (google-workspace-extension.geminicli.com) which governs the agent's authentication state, representing an unverifiable external control-plane dependency with weak/unknown provenance.