account-tiering
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implementation follows secure guidelines for Salesforce integrations, utilizing authorized tools like
dispatch_readonlyfor data access anddisplay_widgetfor presentation. It enforces strict data handling rules, such as grounding custom fields and providing direct links to source records, ensuring transparency and preventing data fabrication. No patterns associated with credential theft, unauthorized code execution, obfuscation, or persistence were found. The connectivity instructions provided are operational best practices for the connector and do not constitute a security bypass. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from Salesforce records and external communication tools (email, Slack, docs), creating a potential surface for indirect prompt injection. 1. Ingestion points: Data retrieved from Salesforce Account, Contact, and Opportunity objects and external signals from email, Slack, and document tools. 2. Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded commands within the ingested data. 3. Capability inventory: The agent has access to Salesforce data tools (dispatch, dispatch_readonly) and a visualization widget that can trigger messaging actions. 4. Sanitization: No explicit content filtering or escaping of ingested data is specified. This finding is classified as safe as it represents the fundamental and expected operation of the skill in analyzing third-party data.
Audit Metadata