delivery-troubleshooter
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill is designed strictly for diagnostic purposes. The instructions explicitly command the agent to perform a READ-ONLY check and include a clear directive not to change budgets, bids, ads, destinations, goals, or statuses. This limits the risk of unauthorized modifications to the user's advertising accounts.
- [NO_CODE]: The skill consists entirely of natural language instructions and configuration metadata. It does not contain any executable scripts, binaries, or external code dependencies, which eliminates the risk of malicious code execution or persistence.
- [DATA_EXPOSURE]: The skill facilitates the analysis of sensitive Google Ads data, such as billing status and conversion performance, via the Model Context Protocol (MCP) or user-provided exports. The instructions focus on processing this data within the session to generate a report and do not contain any patterns suggesting data exfiltration to external servers. The reference to kelpi.ai in the metadata is a legitimate vendor resource linked to the skill's author.
Audit Metadata