vercel-ai-sdk
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches and installs standard libraries from the Vercel AI SDK ecosystem, including "ai", "@ai-sdk/openai", and "@ai-sdk/anthropic", using the project's detected package manager.
- [COMMAND_EXECUTION]: Instructs the agent to run build and test scripts ("npm run build", "npm run test") and install missing dependencies to resolve environment errors autonomously.
- [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: Fetches documentation from "https://ai-sdk.dev/docs/" (SKILL.md).
- Boundary markers: None specified.
- Capability inventory: Can execute bash commands, write/edit files, and fetch web content (Bash, Write, WebFetch).
- Sanitization: No explicit sanitization of fetched documentation before processing.
Audit Metadata