The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill accesses the ~/.google-ads.yaml configuration file, which is a standard repository for sensitive API credentials.
Evidence: SKILL.md contains an instruction to execute cat ~/.google-ads.yaml to verify the setup, which risks exposing raw developer tokens, client secrets, and refresh tokens to the agent's context.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through data ingested from the Google Ads platform.
Ingestion points: The agent reads campaign names, keyword text, and performance metrics from the browser tool (snapshots of ads.google.com/aw/campaigns) and the GoogleAdsService API search results.
Boundary markers: Absent. There are no instructions or delimiters designed to prevent the agent from being influenced by instructions embedded within ad campaign data.
Capability inventory: The skill possesses high-privilege capabilities including pausing campaigns, pausing keywords, and modifying budgets via the API (mutate operations) and browser automation.
Sanitization: Absent. No evidence of input validation or sanitization is present for data retrieved from the external platform before it is used to make operational decisions.
[EXTERNAL_DOWNLOADS]: The skill references the official google-ads Python SDK from PyPI. This is a well-known service provided by Google for API integration.

google-ads