inbound-triggers-monitor
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill performs environment checks and action verification using the official clay CLI tools. These commands are diagnostic in nature and limited to verifying platform availability and workspace identity.
- [SAFE]: The instructions implement a rigorous suppression gate to filter out internal noise, including employees, partners, and company pages, before any paid data enrichment occurs.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from social media interactions. Ingestion points: The social-posts-get-post-interaction-comments tool retrieves raw user words from social media. Boundary markers: No explicit delimiters are used in the processing instructions. Capability inventory: The skill uses clay CLI tools for reading platform data and does not have write or network-send capabilities outside of its primary scope. Sanitization: URL canonicalization is performed on profile links, but text content is not explicitly sanitized.
- [SAFE]: The skill follows security guidelines by advising the agent not to store tokens or passwords in local configuration files.
Audit Metadata