salesforce-hygiene-check
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill uses authorized platform tools (
dispatch_readonly) to interact with the vendor's official APIs (Salesforce GraphQL). It follows least-privilege principles by defaulting to read-only audits and strictly limiting the agent to documented endpoints. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from Salesforce Opportunity records (e.g., NextStep fields) to generate audit suggestions. This presents a potential surface for indirect prompt injection if a record contains malicious instructions. However, the risk is mitigated by the skill's workflow, which mandates human review and manual confirmation before any data is written back to the CRM, preventing autonomous execution of embedded commands.
Audit Metadata