AI & Vector Engineer

What it does

AI & Vector Engineer is a Cursor agent playbook named ai-engineer. AgenticKit installs it so a session can adopt this role instead of acting as a generic coding assistant.

The ai-engineer agent builds production-grade AI features into your SaaS. From streaming chat responses to chunking strategies, semantic search over pgvector/Pinecone, and LLM evaluation, it ensures low latency and high accuracy.

Integrates OpenAI/Anthropic models, embeddings, RAG pipelines, and vector DBs. AI engineer for Cursor specializing in LLM integrations, OpenAI/Anthropic APIs, RAG pipelines, embeddings, prompt engineering, and vector search.

Why it exists

A pile of agent files is not a team. ai-engineer exists so one job, ai & vector engineer, has a written brief, required skills, and a stop condition.

It ships in the Engineering Kit. Skills it must read: LLM Integration & Streaming AI Workflows; Vector Databases & Embeddings Storage; Retrieval-Augmented Generation (RAG) Architecture; System Prompt Engineering & Few-Shot Design. Commands that adopt this role: /ai-feature: AI Integration & Vector Pipeline Builder; /rag-pipeline: RAG Pipeline & Document Ingestion System; /llm-chat: Streaming AI Chat Assistant Interface.

When to use it

• Adding AI chatbots, summarization, or semantic search
• Building document ingestion and retrieval pipelines
• Implementing prompt guards, token cost budgeting, and caching

When not to use it

• Do not ask ai-engineer to do a different role's job. If you need a launch post, switch agents.
• Do not keep the same chat after this agent has finished its artifact. Start a reviewer in a new thread.
• Do not invoke every agent in the kit for a small change.

Example workflow

1. Install the Engineering Kit so .cursor/agents/ai-engineer is on disk.
2. Run /ai-feature: AI Integration & Vector Pipeline Builder, or start a chat and tell Cursor to adopt the ai-engineer role.
3. The agent should read: LLM Integration & Streaming AI Workflows; Vector Databases & Embeddings Storage; Retrieval-Augmented Generation (RAG) Architecture; System Prompt Engineering & Few-Shot Design.
4. It produces the artifact for this role only, then stops.
5. A different agent or you review. Same-chat self-review is not a review.

Example usage

Example: "You are ai-engineer. Adding AI chatbots, summarization, or semantic search. Read llm-integration before you edit."

Or let the pipeline invoke it: /ai-feature: AI Integration & Vector Pipeline Builder.

Capabilities you should actually see: Streaming LLM completions using Vercel AI SDK and Anthropic/OpenAI

Example output

• ai-engineer should leave files or a written verdict, not a vibe check. Streaming LLM completions using Vercel AI SDK and Anthropic/OpenAI Document chunking, embedding generation, and hybrid search RAG
• Engineering agents should touch the slice they were given (schema, route, test, or review note) and nothing else.

Best practices

• Streaming LLM completions using Vercel AI SDK and Anthropic/OpenAI
• Document chunking, embedding generation, and hybrid search RAG
• Prompt engineering with few-shot examples and structured outputs
• Vector search optimization in Pinecone, pgvector, and Qdrant
• One role per chat unless a command is explicitly orchestrating a sequence.

Common mistakes

• Using ai-engineer as a synonym for 'the Cursor agent'. It is a brief, not the product.
• Skipping the required skills and hoping the role name is enough.
• Letting the writer approve its own PR.
• Invoking this agent and three unrelated ones in the same prompt.