solidstart
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides documentation containing code patterns that could introduce vulnerabilities into applications built following these examples.
- Ingestion points: The skill demonstrates ingesting untrusted data through
formDatain API routes (src/routes/api/upload.tsinreferences/server-patterns.md) and through database queries (src/routes/blog/[slug].tsxinSKILL.md). - Boundary markers: No explicit sanitization or validation logic is shown for the data interpolation in the rendering or file writing examples.
- Capability inventory: The skill showcases file system write capabilities using
fs/promises.writeFileinreferences/server-patterns.mdand DOM manipulation viainnerHTMLinSKILL.md. - Sanitization: In
SKILL.md, the usage ofinnerHTML={p().content}lacks sanitization, posing a Cross-Site Scripting (XSS) risk if the content source is untrusted. Inreferences/server-patterns.md, the file upload logic joins a user-provided filename directly to a local path (join(process.cwd(), 'uploads', filename)), creating a Path Traversal vulnerability that could allow attackers to overwrite arbitrary files on the server.
Audit Metadata