customer-success
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE] (SAFE): The skill is entirely instructional and template-based. It describes processes for onboarding, retention, and expansion without requesting any sensitive file access or performing network communications.
- [COMMAND_EXECUTION] (SAFE): While the skill contains a JavaScript code block for a 'Health Score Framework', it is presented as documentation/logic for the agent to follow conceptually rather than a script to be executed in a system shell. There are no shell commands or subprocess calls.
- [DATA_EXFILTRATION] (SAFE): No credentials, API keys, or network-enabled commands were found. The mention of external tools (Salesforce, Gainsight, HubSpot) is for context only and does not include integration code or exfiltration patterns.
- [PROMPT_INJECTION] (SAFE): No instructions to bypass safety filters or override system prompts were detected.
Audit Metadata