miro-automation
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues, malicious instructions, or attempts to bypass safety filters were detected. The skill uses standard Model Context Protocol (MCP) tool patterns.
- [EXTERNAL_DOWNLOADS]: The skill references an external MCP server endpoint at https://rube.app/mcp. This is a standard configuration step for connecting to third-party toolkits via MCP and does not involve unverifiable or dangerous code execution.
- [PROMPT_INJECTION]: The skill processes user-generated content from Miro boards, which presents a surface for indirect prompt injection. (Ingestion points: MIRO_GET_BOARDS2 and MIRO_GET_BOARD_ITEMS retrieve board metadata and item content; Boundary markers: None specified in instructions; Capability inventory: Includes the MIRO_SHARE_BOARD tool which could be misused if instructions are followed; Sanitization: No sanitization logic for retrieved content is described).
Audit Metadata