data360-query
Pass
Audited by Gen Agent Trust Hub on Aug 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands through the Salesforce CLI (
sf data360) and local Node.js scripts (diagnose-org.mjs) to perform queries, describe tables, and manage search indexes. These are standard operations for the skill's purpose. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes and displays data retrieved from external sources (Salesforce Data Cloud).
- Ingestion points: Data enters the agent's context through SQL query results (
sf data360 query sql), vector search results (sf data360 query vector), and hybrid search results (sf data360 query hybrid) as documented inSKILL.md. - Boundary markers: The skill lacks explicit instructions or delimiters to tell the agent to ignore potentially malicious instructions embedded within the query results.
- Capability inventory: The agent has the capability to execute shell commands using the
sfCLI andnodeas defined in theSKILL.mdworkflow. - Sanitization: There is no evidence of sanitization, filtering, or validation of the retrieved data before it is processed by the agent.
Audit Metadata