miro-automation

Pass

Audited by Gen Agent Trust Hub on Oct 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill utilizes the Rube MCP service via https://rube.app/mcp to retrieve tool schemas and manage Miro connections. This represents the core infrastructure required for the skill's stated purpose of Miro automation.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from Miro boards, creating a potential surface for indirect prompt injection where instructions embedded in board items could influence agent behavior.
  • Ingestion points: The skill reads board content and metadata through tools such as MIRO_GET_BOARD_ITEMS, MIRO_GET_BOARDS2, MIRO_GET_BOARD, and MIRO_GET_CONNECTORS2 as documented in SKILL.md.
  • Boundary markers: There are no instructions or delimiters defined to help the agent distinguish between administrative instructions and the data retrieved from Miro boards.
  • Capability inventory: The skill includes powerful capabilities such as sharing boards with new users (MIRO_SHARE_BOARD), creating new items (MIRO_CREATE_STICKY_NOTE_ITEM), and bulk creation of content (MIRO_CREATE_ITEMS_IN_BULK), which could be misused if the agent obeys instructions found within a board.
  • Sanitization: The instructions do not specify any validation or sanitization of the content retrieved from Miro boards before it is used by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 8, 2026, 09:50 PM