salesforce

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data retrieved from Salesforce (e.g., Contact and Opportunity records), which introduces a surface for indirect prompt injection if that data contains malicious instructions.\n
  • Ingestion points: Salesforce API responses retrieved via gateway.maton.ai as documented in SKILL.md.\n
  • Boundary markers: The provided code examples do not include explicit delimiters or instructions to ignore embedded commands within the CRM data.\n
  • Capability inventory: The skill utilizes network capabilities via Python libraries (urllib, requests) to interact with external endpoints.\n
  • Sanitization: No data validation or sanitization routines are demonstrated for the content retrieved from the external API.\n- [DATA_EXFILTRATION]: The skill communicates with gateway.maton.ai and ctrl.maton.ai. These domains represent vendor-managed infrastructure for the skill's API gateway and connection management services.\n- [COMMAND_EXECUTION]: Documentation includes usage examples that execute Python code via shell here-docs (python <<'EOF'), which involves piping multi-line scripts to an interpreter.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 01:52 PM
Security Audit — agent-trust-hub — salesforce