storyblok-model-content
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from external URLs, design sketches, or written briefs to generate content models and schema updates.
- Ingestion points: SKILL.md (Step 1) specifies fetching page content from a user-provided URL or reading requirements from external briefs and images.
- Capability inventory: The skill possesses significant capabilities, including the ability to perform destructive schema deletions and modifications via API calls (createComponent, updateComponent, deleteComponent) or CLI commands (schema push).
- Boundary markers: While the skill mandates a structured extraction process and a 'plan-confirm-execute' gate, it does not specify explicit delimiters or 'ignore' instructions for the processed content.
- Sanitization: No specific sanitization or validation of the fetched external text is described prior to its use in the modeling process.
- [COMMAND_EXECUTION]: The skill instructions involve executing shell commands via the Storyblok CLI in Code mode, such as 'storyblok schema push' and 'storyblok schema validate' as referenced in resources/schema-as-code.md.
- [DATA_EXFILTRATION]: The skill performs network operations to fetch external web pages provided by the user (Step 1). While this is intended for modeling requirements, fetching data from arbitrary non-whitelisted domains represents an outbound network surface.
Audit Metadata