storybook-skills-standard
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow that ingests untrusted data from a local project environment, creating a potential surface for indirect prompt injection.\n
- Ingestion points: As part of establishing project context, the agent is instructed to read
package.json,.storybook/main.*, and existing story files within the user's workspace.\n - Boundary markers: The skill does not provide instructions to use delimiters or ignore embedded instructions when reading these local project files, increasing the risk of the agent following instructions contained within the analyzed code instead of the skill's own rules.\n
- Capability inventory: The agent is directed to execute existing Storybook, test, lint, or build commands defined in the local repository's scripts, which provides an execution path for potentially malicious code defined in the workspace.\n
- Sanitization: There are no explicit requirements for the agent to validate or sanitize project-provided data or script names before they are used in the workflow.
Audit Metadata