figma-integration
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill involves the execution of a local script,
scripts/validate_tokens.py, to verify token integrity after synchronization. This is a standard and intended project workflow for maintaining data consistency. - [PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection due to its core function of ingesting data from external design assets.
- Ingestion points: The agent is instructed to read frames, variables, and screenshots from Figma via the REST API or an MCP server (as noted in SKILL.md).
- Boundary markers: There are no explicit instructions or delimiters provided to the agent to treat external Figma content as untrusted or to ignore embedded instructions within that data.
- Capability inventory: The agent possesses the capability to execute shell commands (running the validation script) and perform network operations through the Figma MCP tools.
- Sanitization: No evidence of data sanitization, filtering, or validation is mentioned before the external design content is interpolated into the agent's working context.
Audit Metadata