gsc-portfolio-audit

Pass

Audited by Gen Agent Trust Hub on Aug 7, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill uses the 'requests' library to interact with Google Search Console APIs.
  • [DATA_EXFILTRATION]: Communicates search performance data from official Google endpoints to the local environment. No data is sent to unauthorized or untrusted third-party domains.
  • [CREDENTIALS_UNSAFE]: Utilizes Google OAuth credentials (Client ID, Secret, and Refresh Token) sourced from environment variables. This follows standard secure practices for API authentication.
  • [PROMPT_INJECTION]: Potential surface for indirect prompt injection as the tool ingests search query data from an external source (Google Search Console API).
  • Ingestion points: Fetches raw search queries and performance metrics in 'scripts/audit_gsc.py'.
  • Boundary markers: No delimiters or isolation prompts are used when displaying query data to the agent.
  • Capability inventory: The script is limited to API requests and console output; it lacks dangerous capabilities like arbitrary file writing or shell command execution.
  • Sanitization: Search query strings are processed and displayed without sanitization or filtering.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 7, 2026, 06:18 PM
Security Audit — agent-trust-hub — gsc-portfolio-audit