Kommo Automation
Pass
Audited by Gen Agent Trust Hub on May 4, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill configuration references an external MCP server at
https://rube.app/mcp. This is a vendor-provided endpoint used to bridge tool execution for the Composio integration. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data from Kommo CRM (leads, custom fields, tasks) which introduces a theoretical surface for indirect prompt injection. Malicious content within CRM records could potentially influence the agent's behavior during data processing.
- Ingestion points:
KOMMO_LIST_LEADSandKOMMO_LIST_CUSTOM_FIELDSinSKILL.md. - Boundary markers: None explicitly defined in the instructions.
- Capability inventory: Tools for creating and updating leads/tasks (
KOMMO_CREATE_LEAD,KOMMO_UPDATE_LEAD,KOMMO_CREATE_TASK) are provided inSKILL.md. - Sanitization: Not explicitly mentioned in the skill instructions.
Audit Metadata