sales-account-segmentation

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructions describe a workflow that is susceptible to indirect prompt injection due to the ingestion of untrusted external data.
  • Ingestion points: In SKILL.md, the instructions explicitly direct the agent to "read CRM exports or the company's website" to pre-fill interview answers and draft candidate models. It also involves processing "product-usage data" and "web analytics."
  • Boundary markers: The instructions do not specify the use of delimiters or provide the agent with "ignore embedded instructions" warnings for the external data it is expected to process.
  • Capability inventory: The skill generates a structured "model charter" as text output and utilizes the platform's "persistent memory" to store decisions. It does not perform arbitrary command execution or network exfiltration.
  • Sanitization: There are no instructions for sanitizing or validating the content retrieved from CRM exports or websites before it is incorporated into the agent's reasoning process.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 09:18 PM
Security Audit — agent-trust-hub — sales-account-segmentation