salesforce
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFE
Full Analysis
- Data Ingestion Surface: The skill is designed to retrieve information from Salesforce records, including activity history, account plans, and conversation transcripts. As with any system that processes external data, there is a potential surface for indirect prompt injection if the source records contain instructions. However, the skill mitigates this by requiring metadata verification and structured SOQL queries rather than broad text searches.
- Controlled Capability Inventory: The skill explicitly restricts the agent to a specific set of Salesforce tools (e.g.,
describe_sobject,soql_query,create_record). It provides clear guidance on write safety and warns against using unexposed surfaces like SOSL or Bulk API, adhering to the principle of least privilege. - Preflight Context Dependency: The skill references a mandatory pre-answer gate (
sales:user-context) located in a relative path. This is a standard architectural pattern for managing user session state and permissions within the agent environment. - Write Safety Protocols: The instructions include mandatory checks before performing data mutations, such as confirming record IDs and field API names, and verifying outcomes after updates. This reduces the risk of accidental or unauthorized data changes.
Audit Metadata