apple-ads
Pass
Audited by Gen Agent Trust Hub on Aug 18, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill ingests and displays search term text from the Apple Ads API, which is an external source that could be influenced by third-party search queries. This data is processed by the AI agent without explicit delimiters, creating a surface for indirect prompt injection.\n
- Ingestion points:
scripts/fetch-ads.tsfetchessearchTermTextfrom the Apple Ads Platform API and formats it for the agent's context.\n - Boundary markers: Data is presented in Markdown tables but lacks explicit delimiters or instructions to ignore embedded content.\n
- Capability inventory: The skill has significant capabilities to modify ad account state via
scripts/apply-change.ts.\n - Sanitization: No sanitization or validation of the search term content is performed before presentation to the agent.\n- [COMMAND_EXECUTION]: The skill relies on local script execution (
scripts/fetch-ads.ts,scripts/apply-change.ts,scripts/inspect-account.ts) via thebunruntime to manage the Apple Ads account. These operations are essential to the skill's purpose and are intended for use with human-in-the-loop oversight as described in the documentation.
Audit Metadata