What problem does it solve? Adding long-term memory to LLM applications built with the Vercel AI SDK requires manual retrieval, prompt injection, and storage logic. This Skill guides Claude to set up @mem0/vercel-ai-provider so memories are automatically retrieved and stored during generateText and streamText calls. ## Core Features & Use Cases - Wrapped Model Setup: Configure createMem0 to wrap OpenAI, Anthropic, Google, Groq, or Cohere models with automatic memory retrieval and storage. - Standalone Memory Utilities: Generate code using retrieveMemories, getMemories, searchMemories, and addMemories for full manual control over the memory lifecycle. - Framework Integration: Produce working examples for Next.js API routes, streaming responses, structured output with generateObject, and multi-provider configurations. - Use Case: A developer building a Next.js chatbot asks Claude to add per-user memory; the Skill produces an API route that retrieves relevant memories, injects them as a system prompt, streams the response, and stores the conversation back to Mem0. ## Quick Start Ask Claude to add Mem0 memory to your Vercel AI SDK app, for example: set up mem0 with streamText in my Next.js API route.
