datahub-mfe-create-app
Pass
Audited by Gen Agent Trust Hub on Apr 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands such as
npm installandnpm startwithin the newly created application directory. - [EXTERNAL_DOWNLOADS]: Dependencies are fetched from the public npm registry during the
npm installprocess. The requested packages are standard, well-known libraries for React and Webpack development. - [REMOTE_CODE_EXECUTION]: The skill generates project configuration and source files using templates and user-provided inputs, then executes the resulting scripts through the npm CLI. This constitutes dynamic execution of code generated at runtime.
- [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection. Ingestion points: User input collected via
AskQuestioninSKILL.md. Boundary markers: Absent. Capability inventory: Performs file writes and shell execution (npm install,npm start) as described inSKILL.md. Sanitization: Absent; user input is interpolated directly into templates.
Audit Metadata