google-ads-query
Pass
Audited by Gen Agent Trust Hub on Jul 30, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill processes data from the Google Ads API, which constitutes an external data ingestion surface. It mitigates indirect prompt injection risks by using a 'CSV-first' pattern where raw data is stored in local files and only minimal metadata (path and row count) is returned to the context. (Ingestion points: Google Ads API via query.py; Boundary markers: Explicit analysis request required before reading files; Capability inventory: Read, Bash, and Glob tools; Sanitization: Results are flattened into structured CSV data).
- [CREDENTIALS_UNSAFE]: The skill requires a google-ads.yaml file for authentication. It follows security best practices by instructing users to maintain this file locally and outside of version control, ensuring sensitive API keys are not exposed.
- [EXTERNAL_DOWNLOADS]: Core scripts and templates are retrieved from the author's verified GitHub repository (fourteenwm/ppc-ai-skills). These resources are transparently documented as part of the installation process and are essential to the skill's primary function.
- [COMMAND_EXECUTION]: The skill's Python logic is strictly limited to read-only SELECT queries through the Google Ads Service. It does not contain functionality for data mutation, administrative actions, or the execution of arbitrary shell commands.
Audit Metadata