ui-convert-writer-stitch
Pass
Audited by Gen Agent Trust Hub on Apr 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted Intermediate Representation (IR) files to drive its logic and MCP tool execution. If these files are sourced from an external or malicious actor, they could contain instructions designed to manipulate the agent's output or misuse the available design tools.
- Ingestion points: Reads IR JSON files from the
.ui-convert/ir/directory. - Boundary markers: No specific delimiters or safety instructions are defined to separate IR data from instructions.
- Capability inventory: Accesses a wide array of MCP tools including
create_frame,create_text,create_component, andset_tokenswhich allow modification of the target design project. - Sanitization: The instructions do not specify any validation or sanitization of the IR data before it is mapped to MCP tool parameters.
Audit Metadata