stitch-vue-vant-components
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads design metadata, screenshots, and HTML content from Google Cloud Storage URLs provided by the Stitch MCP tool. These downloads are necessary for the skill's primary function and target well-known infrastructure.
- [COMMAND_EXECUTION]: Executes a local bash script (scripts/fetch-stitch.sh) to perform asset downloads via curl. This script handles network-level details like redirects and timeouts for reliability.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes external HTML data to generate source code. (1) Ingestion points: External HTML content retrieved from design URLs provided by the Stitch MCP tool. (2) Boundary markers: No specific delimiters or instructions are used to wall off external content from the code generation logic. (3) Capability inventory: The skill utilizes file-writing (Write) and shell execution (Bash) capabilities to create the project structure. (4) Sanitization: There is no evidence of HTML sanitization or instruction filtering before the content is used to draft Vue components.
Audit Metadata