ai-sdk
Vercel AI SDK expert guidance. Use when building AI-powered features — chat interfaces, text generation, structured output, tool calling, agents, MCP integration, streaming, embeddings, reranking, image generation, or working with any LLM provider.
76
74%
Does it follow best practices?
Impact
76%
6.33xAverage score across 3 eval scenarios
Low
Low-risk findings worth noting
Fix and improve this skill with Tessl
tessl review fix ./skills/ai-sdk/SKILL.mdEvaluation results
96%
82%Customer Support Agent for an E-Commerce Platform
Type-safe ToolLoopAgent customer support bot with AI SDK v6 conventions
ToolLoopAgent class used
0%
100%
agent.generate() or agent.stream()
0%
100%
stopWhen: stepCountIs(N)
0%
100%
inputSchema in tool definitions
0%
100%
Tool files in lib/tools/
0%
100%
Agent file in lib/agents/
0%
100%
InferAgentUIMessage exported
0%
100%
uiMessages property in API route
0%
100%
Gateway model format
0%
100%
No direct provider SDK imports
100%
100%
ModelMessage / convertToModelMessages
0%
50%
Typecheck output saved
100%
100%
47%
32%Product Catalog Digitization Tool
Structured product extraction with AI SDK v6 APIs
generateText import
0%
100%
Output import
0%
0%
Output.array or Output.object used
0%
0%
No generateObject
100%
100%
Zod schema defined
0%
0%
AI Gateway model format
0%
0%
No direct provider SDK import
0%
100%
Runtime model ID fetch
0%
40%
maxOutputTokens for token limit
0%
100%
Typecheck evidence
100%
80%
86%
79%Real-Time Weather Chat Interface
Next.js streaming chat with weather tool (AI SDK v6 APIs)
Stream response method
0%
70%
useChat transport
0%
0%
useState input state
100%
100%
sendMessage submission
0%
100%
message.parts rendering
0%
100%
Typed tool part name
0%
100%
Tool part data properties
0%
100%
Tool part state guards
0%
100%
status not isLoading
0%
100%
AI Gateway model
0%
100%
inputSchema in tool
0%
100%
stopWhen stepCountIs
0%
88%
- Repository
- vercel/vercel-plugin
- Commit
358400a
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.