ai-sdk

Pass

Audited by Gen Agent Trust Hub on Jul 11, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill instructs the agent to disregard its training data regarding the AI SDK, citing that internal knowledge may be obsolete. It directs the agent to prioritize information found in current documentation and source code to ensure technical accuracy.- [EXTERNAL_DOWNLOADS]: The skill directs the installation of the ai package and related libraries (e.g., @ai-sdk/openai, @ai-sdk/react) from standard registries. It also fetches model metadata from Vercel's official gateway.- [COMMAND_EXECUTION]: The skill utilizes curl to retrieve the latest model identifiers from ai-gateway.vercel.sh and employs grep for searching API definitions within the local node_modules directory.- [REMOTE_CODE_EXECUTION]: The skill references the use of npx @ai-sdk/devtools for launching a local development interface to inspect and debug AI SDK operations.- [DATA_INJECTION_SURFACE]: The skill processes documentation content from local files and the ai-sdk.dev website to guide code generation.
  • Ingestion points: grep searches on node_modules/ai/docs/ and web requests for documentation at https://ai-sdk.dev/docs/.
  • Boundary markers: None; documentation content is processed directly to formulate responses.
  • Capability inventory: Includes package management (pnpm), network operations (curl), and utility execution (npx).
  • Sanitization: Documentation content is used for technical reference without specific sanitization routines.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 11, 2026, 07:04 PM