google-auth
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill provides templates for processing external authentication tokens, which serves as an indirect prompt injection surface.
- Ingestion points: Authentication codes and ID tokens are ingested via web callback handlers and verification functions described in references/oauth2-web-server.md and references/sign-in-with-google.md.
- Boundary markers: The documentation instructs developers to verify JWT signatures, audience, and issuer claims using official library methods.
- Capability inventory: The provided templates enable authenticated network API calls through official Google libraries.
- Sanitization: Examples include CSRF protection using state parameters and double-submit-cookie validation for g_csrf_token.
- [COMMAND_EXECUTION]: The documentation describes the usage of the executable-sourced credentials feature within Google's authentication framework.
- Evidence: references/workload-identity.md provides guidance on configuring the --executable-command parameter and enabling the GOOGLE_EXTERNAL_ACCOUNT_ALLOW_EXECUTABLES environment variable.
- Context: The skill includes specific security considerations, recommending that users restrict access to configuration files and protect against unauthorized process execution.
Audit Metadata