stitch-uview-plus-components
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes a shell script
scripts/fetch-stitch.shto download design assets. This script executescurlwith parameters provided by the agent at runtime. - [EXTERNAL_DOWNLOADS]: The skill fetches HTML and visual assets from
stitch.withgoogle.com. These downloads are part of the intended workflow to retrieve design specifications from a well-known Google platform. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted HTML and metadata exported from external design projects. This content is used as a template for generating Vue.js components, creating a potential surface for instructions embedded in the source data to influence the agent's output.
- Ingestion points: Design HTML fetched via
htmlCode.downloadUrland saved totemp/source.html. - Boundary markers: No explicit delimiters or instructions are provided to the agent to disregard potentially malicious content within the source HTML.
- Capability inventory: The skill can execute shell commands for fetching data and performs extensive filesystem write operations to generate components.
- Sanitization: The skill lacks explicit sanitization or validation steps for the external HTML before it is parsed and used for code generation.
Audit Metadata