next-theme-figma
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill's test suite (located in the tests/ directory) utilizes subprocess.run to execute its own local Node.js and Python utility scripts (scripts/theme-figma.js and scripts/copy-lint.py) for functional verification. These executions are scoped to the local environment and are used to validate generated handoff packages.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external Figma designs (text layers, metadata, and variables). This data is processed by the agent to generate handoff manifests, creating a potential vector for indirect prompt injection where malicious instructions in a design file could attempt to manipulate agent behavior. Mandatory Evidence: 1. Ingestion points: Design text, variables, and metadata fetched from Figma via tool calls. 2. Boundary markers: The skill instructions emphasize the use of structured manifests and a 'Platform Divergence Ledger' to separate design intent from platform code. 3. Capability inventory: Local file write access for manifest generation and Bash execution for running helper scripts. 4. Sanitization: Handoff package files are strictly validated against pre-defined JSON schemas by the internal theme-figma.js script.
Audit Metadata