ai-sdk
Pass
Audited by Gen Agent Trust Hub on May 1, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill contains directives intended to override the agent's internal knowledge base, such as "Everything you know about the AI SDK is outdated or wrong" and "Never rely on memory."
- [PROMPT_INJECTION]: The skill identifies an indirect prompt injection surface as it is instructed to ingest and process external markdown documentation from the web. • Ingestion points: Remote markdown files from
ai-sdk.dev. • Boundary markers: Absent for external documentation. • Capability inventory: Shell command execution (pnpm,curl,npx) and local file system access. • Sanitization: No sanitization or validation of external content is specified. - [COMMAND_EXECUTION]: The skill instructs the agent to execute several shell commands for development and environment setup:
pnpm add aifor package installation,grepfor searching within thenode_modulesdirectory,curlfor retrieving model metadata fromai-gateway.vercel.sh, andnpx @ai-sdk/devtoolsfor running development tools. - [EXTERNAL_DOWNLOADS]: The skill fetches model configuration and documentation from
ai-gateway.vercel.shandai-sdk.dev. These domains are associated with Vercel's service infrastructure.
Audit Metadata