local-business-marketing
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implements robust security and operational guardrails, requiring the agent to verify all business-critical information (hours, locations, offer terms) and obtain explicit user approval before performing any actions that affect external accounts.
- [SAFE]: The skill defines a dependency on an external Model Context Protocol (MCP) tool hosted by a well-known marketing service provider. This integration is used as intended for social media workflows, with the skill's instructions ensuring all automated actions are initiated as drafts for user review.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external data sources such as business FAQs, customer reviews, and local event information. While such data could theoretically contain malicious instructions, the skill's mandatory requirement for human review and explicit approval of all drafts before publication serves as an effective control to prevent the execution of injected content.
Audit Metadata