deal-momentum-analyzer

Warn

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: MEDIUMPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [PERSISTENCE]: The skill documentation describes methods to maintain execution across sessions for daily automation. Specifically, it suggests using the CronCreate tool or creating a persistent macOS launchd agent at ~/Library/LaunchAgents/com.tim.deal-momentum.plist to ensure the analysis runs every weekday at 7:00 AM CST.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes data from external, potentially untrusted sources such as CRM notes and call transcripts, which could be manipulated to influence agent behavior.
  • Ingestion points: Untrusted data enters the agent context through clari_get_call_summary, hubspot_search_contacts, and the ask_agent tool used to pull activity history.
  • Boundary markers: The instructions do not specify the use of delimiters or protective instructions to ignore potentially malicious content embedded within CRM notes or transcripts.
  • Capability inventory: The skill possesses capabilities to create Gmail drafts, generate calendar events, and execute Apollo people searches based on the analyzed data.
  • Sanitization: There is no evidence of sanitization or validation of the ingested text before it is used to determine next-best-actions or generate email content.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 17, 2026, 09:56 AM
Security Audit — agent-trust-hub — deal-momentum-analyzer