dt-figma-export

Pass

Audited by Gen Agent Trust Hub on Oct 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • Interaction with Well-Known Services: The skill integrates with Figma via a Model Context Protocol (MCP) server. This involves sending project artifacts to Figma for visualization, which is the primary intended purpose of the skill. The skill provides clear setup instructions for this official integration.
  • Data Ingestion and Processing: The skill reads local files from the .copilot-tracking/dt/ directory. While this provides a surface for potential indirect prompt injection if the artifacts were to contain instructions, the skill includes explicit constraints to treat all ingested payloads as data rather than instructions, which is a standard safety measure.
  • User Confirmation for Write Operations: The skill implements a robust safety control by requiring the agent to present the exact destination and operation to the user and obtain explicit confirmation before calling any write-capable tools like figma/use_figma or figma/create_new_file.
  • Reference Implementation Templates: The skill provides JavaScript snippets for the Figma Plugin API in its reference documentation. These serve as structural guides for the agent when formatting Figma elements and do not involve local shell execution or unauthorized code running on the host system.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 8, 2026, 10:57 AM