zoho_desk-automation

Fail

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [Indirect Prompt Injection] (HIGH): The skill is highly vulnerable to indirect prompt injection as it processes external content from Zoho Desk (tickets, comments, and contacts).
  • Ingestion points: Data fetched via RUBE_SEARCH_TOOLS and RUBE_MULTI_EXECUTE_TOOL from Zoho Desk tickets and customer interactions.
  • Boundary markers: None identified; instructions do not specify any delimiters to separate untrusted data from system instructions.
  • Capability inventory: The agent has write/modify access through RUBE_MULTI_EXECUTE_TOOL and orchestrated execution capabilities via RUBE_REMOTE_WORKBENCH.
  • Sanitization: No sanitization or validation of the ingested ticket content is mentioned before it is processed by the agent.
  • [Unverifiable Dependencies] (MEDIUM): The skill requires the user to add https://rube.app/mcp as an MCP server. This is an external service not listed in the trusted repositories or organizations, creating a dependency on an unverified third-party platform that intermediates help desk data.
  • [Dynamic Execution] (MEDIUM): The use of RUBE_REMOTE_WORKBENCH for "bulk operations or data processing" suggests a remote runtime environment where the agent can execute code (e.g., Python loops with ThreadPoolExecutor). While functional, this provides a powerful execution surface that could be abused if the agent is influenced by malicious ticket data.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 16, 2026, 09:45 AM