cocos-ui-schema-generator
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions in
SKILL.mddirect the agent to run a PowerShell script (scripts/install-runtime.ps1) using the-ExecutionPolicy Bypassflag. This command is intended to copy the bundled WebUI runtime files into a target Cocos project directory during the initial setup phase. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data, such as HTML, UI mockups, and textual descriptions, to generate code, which represents an indirect injection surface.
- Ingestion points: As described in
SKILL.md, the agent accepts screenshots, HTML, mockups, and textual descriptions as the primary inputs for generating UI schemas. - Boundary markers: The instructions lack specific guidance on using delimiters (such as XML tags or triple backticks) to separate the untrusted user-provided UI descriptions from the agent's core instructions.
- Capability inventory: The skill possesses the ability to perform file system operations (creating
.tsschema files) and execute local scripts for installation and validation. - Sanitization: No explicit instructions are provided for the agent to sanitize or filter the content of the UI descriptions to prevent embedded instructions from influencing its behavior.
Audit Metadata