ui-workflow
Warn
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: MEDIUMPRIVILEGE_ESCALATIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PRIVILEGE_ESCALATION]: The skill directs the agent to "request escalated permissions for Storybook CLI commands" if the execution environment blocks network access or port binding, which encourages bypassing standard sandbox restrictions.
- [EXTERNAL_DOWNLOADS]: The instructions utilize
npx storybook, which downloads the Storybook package and its associated dependencies from the npm registry during execution. - [REMOTE_CODE_EXECUTION]: The command
npx storybook skills ui-workflowexecutes code from an external package within the agent's environment. - [INDIRECT_PROMPT_INJECTION]: The skill mandates that the agent follow the output of an external CLI command as a "mandatory, ordered workflow." This creates a vulnerability where dynamically generated content from a tool can influence or override the agent's instruction set at runtime.
- [COMMAND_EXECUTION]: The skill requires the agent to run several shell commands, including starting a background development server using the project's package manager and executing Storybook-specific CLI utilities.
Audit Metadata