skills/jeremylongshore/tons-of-skills-marketplace/salesforce-migration-deep-dive/Gen Agent Trust Hub
salesforce-migration-deep-dive
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by ingesting and processing data from external Salesforce environments.
- Ingestion points: Data is retrieved from Salesforce orgs via SOQL queries (
conn.query) and API requests (conn.request) inSKILL.md. - Boundary markers: There are no explicit delimiters or instructions to ignore embedded malicious content within the retrieved Salesforce data.
- Capability inventory: The skill possesses significant capabilities including the ability to write to the file system, execute bash commands, and perform destructive operations in Salesforce (upserting data, deleting records via
rollbackMigration). - Sanitization: The TypeScript snippets use template literals to construct SOQL queries (e.g.,
SELECT COUNT(Id) total FROM ${obj}), which could lead to SOQL injection if the object names or retrieved data were influenced by an attacker and not properly validated before interpolation.
Audit Metadata