account-plan

Pass

Audited by Gen Agent Trust Hub on Nov 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill uses official Salesforce GraphQL and REST APIs (v63.0, v65.0, v66.0) for all data operations. All endpoints and tools (dispatch, dispatch_readonly) are standard Salesforce connector capabilities.
  • [COMMAND_EXECUTION]: The skill uses a dynamic context injection pattern (dispatch_readonly(...)) to verify server connectivity. This is used for a legitimate connectivity check and does not involve exfiltration or execution of untrusted input.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from Salesforce Account and Opportunity records. It mitigates risk by requiring explicit user confirmation (the 'write gate') before creating or updating records based on that data, and uses markers like [needs validation] for inferred content.
  • [DATA_EXFILTRATION]: No exfiltration detected. While the skill performs web research to ground account data, it does not transmit sensitive Salesforce data to untrusted third-party domains. Salesforce record links are constructed using the user's own instance URL.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 11, 2026, 06:39 PM