geistdocs
Pass
Audited by Gen Agent Trust Hub on Mar 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [Command Execution]: The skill references
npx @vercel/geistdocs initfor project initialization. This is a standard procedure for bootstrapping Vercel projects and uses official vendor tools. - [External Dependencies]: The skill utilizes the
@vercel/geistdocsandfumadocsecosystem, which are core components of the documentation framework and originate from recognized sources. - [Indirect Prompt Injection Surface]: The 'Ask AI' feature processes content from documentation files to answer queries, which is a potential surface for indirect prompt injection.
- Ingestion points: Untrusted documentation content is ingested from MDX files within the
content/docs/directory as defined in the skill metadata. - Boundary markers: The skill relies on standard MDX frontmatter and Markdown structure to delineate content boundaries.
- Capability inventory: The integrated AI assistant utilizes a
search_docstool and supports user file/image uploads for expanded context. - Sanitization: Input validation and safety protocols are facilitated through the Vercel AI Gateway specified in the deployment requirements.
Audit Metadata