VoltAgent

VoltAgent is an AI Agent Engineering Platform for building and operating AI agents in TypeScript. It is built for developers who want full code control over agent logic while still getting production-ready visibility and operations, combining an open-source framework with a managed or self-hosted console. The framework covers memory, RAG, guardrails, tools, MCP, voice and workflows.

A developer scaffolds a project with the create-voltagent-app CLI, which generates a starter agent and a sample workflow. Agents are defined in code with a name, instructions, a model provider and tools, and multiple specialized agents can run under a supervisor that routes tasks between them; the VoltOps Console then shows live chat, traces and logs for the running agents.

VoltAgent's workflow engine describes multi-step automations declaratively, including human-in-the-loop steps that suspend and resume for manager approval. It swaps between OpenAI, Anthropic, Google and other providers without rewriting agent logic, attaches durable memory across runs, grounds responses with retriever agents and RAG, and adds text-to-speech or speech-to-text through OpenAI, ElevenLabs or custom voice providers, with guardrails to validate input and output at runtime.

VoltAgent's framework is released under the MIT license, installs via npm and is built around Node.js packages such as @voltagent/core. The VoltOps Console is available as a managed cloud service or can be self-hosted, and an MCP server (@voltagent/mcp-docs-server) lets AI coding assistants such as Claude, Cursor or Windsurf access VoltAgent's documentation directly while coding.

Use cases

• Building a multi-agent system where specialized agents work under a supervisor
• Automating an expense approval workflow with manager approval for high amounts
• Teaching an AI coding assistant to use VoltAgent via its MCP docs server