Follow-Up Reply Drafter
Pass
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill is entirely instructional and does not contain any executable scripts, code, or binary files.
- [SAFE]: References to well-known technology organizations and services such as HubSpot, Calendly, and Gong.io are used neutrally to provide legitimate industry benchmarks and guidelines.
- [PROMPT_INJECTION]: The skill processes untrusted email threads which creates an indirect prompt injection surface. 1. Ingestion points: The 'context' input and email data retrieved via 'search_emails'. 2. Boundary markers: No specific delimiters are present in the drafting instructions. 3. Capability inventory: Read-only access to email and CRM data via platform tool calls. 4. Sanitization: No explicit content validation is described. This surface is limited to text generation and does not execute sensitive operations.
- [SAFE]: No evidence of obfuscation, data exfiltration, or unauthorized command execution was found in the skill metadata or reference files.
Audit Metadata