claroty-risk-triage

Pass

Audited by Gen Agent Trust Hub on Jul 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local Python script located at mcp-servers/claroty-mcp/claroty_mcp_server.py to run the Model Context Protocol server.
  • [PROMPT_INJECTION]: The skill processes untrusted data from Claroty xDome APIs, creating a potential surface for indirect prompt injection. 1. Ingestion points: Alerts and vulnerability details fetched via list and get tools. 2. Boundary markers: Not specified in the instruction set. 3. Capability inventory: Includes write operations such as alert acknowledgement, labeling, and assignment. 4. Sanitization: Not explicitly documented in the provided file.
  • [SAFE]: Authentication tokens and API endpoints are managed securely via environment variables (CLAROTY_API_TOKEN, CLAROTY_API_URL). The skill also enforces an ITSM gating mechanism for write operations to ensure authorized changes.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 30, 2026, 04:29 PM
Security Audit — agent-trust-hub — claroty-risk-triage