figma-fetch-reconcile-and-analyse
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill facilitates legitimate design-to-development workflows by extracting metadata from Figma nodes via a Model Context Protocol (MCP) server.
- [SAFE]: No evidence of unauthorized data access or exfiltration. The skill reads from local project files (Jira output) and writes to specific source directories (.src/figma-output/) without accessing sensitive user credentials or system paths.
- [SAFE]: There are no indicators of prompt injection, obfuscation, or remote code execution. Instructions are clearly defined and focused on data transformation and structural analysis.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from Jira stories and Figma nodes. (1) Ingestion points: reads .SS_WF/{{$var[ticket_id]s}}_jira_output.json and Figma node metadata via MCP tools. (2) Boundary markers: None explicitly defined to separate external data from instructions. (3) Capability inventory: writes files to .src/figma-output/ and invokes Figma MCP tool calls. (4) Sanitization: No explicit sanitization or validation of the ingested text is mentioned. However, the risk is minimized as the skill is instructed to extract structured metadata (dimensions, tokens, hierarchy) rather than interpreting the content as instructions.
Audit Metadata