cms
Fail
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: CRITICALEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- Placeholder URL References: The skill includes a placeholder URL (https://your-site.vercel.app/api/revalidate) for webhook configuration. While generic subdomains can be flagged by automated reputation systems, its usage here is strictly illustrative for developer configuration purposes.\n- External Package Integration: The skill suggests installing standard libraries such as next-sanity, contentful, and @builder.io/sdk from public registries. These are official tools provided by the CMS vendors for platform integration.\n- Indirect Prompt Injection Surface: The skill facilitates fetching data from external CMS platforms, which serves as a potential entry point for untrusted content.\n
- Ingestion points: External data is retrieved via sanityFetch and contentful.getEntries (seen in app/page.tsx).\n
- Boundary markers: No specific prompt boundaries are implemented in the rendering examples.\n
- Capability inventory: The skill utilizes cache revalidation (revalidateTag) and navigation (Response.redirect) capabilities.\n
- Sanitization: Examples show direct rendering of content (e.g., {post.title}) without explicit sanitization in the snippets.\n
- Context: Security is maintained through isValidSignature checks on webhooks and secret-based authentication for Draft Mode, ensuring only authorized sources can influence application state.\n- CLI-Based Configuration: It recommends using vercel integration add for provisioning environment variables, which is a standard method for managing sensitive tokens and project IDs within the Vercel ecosystem.
Recommendations
- Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata