 1
 2          const memory = new Memory({
 3  storage: new LibSQLStore({
 4    url: "file:../mastra.db",
 5  }),
 6});

 1
 2const copywriterAgent = new Agent({
 3  instructions,
 4  model: anthropic("claude-3-5-sonnet"),
 5  memory
 6});

Mastra Agents

Stateful AI agents with any LLM.

Mastra agents use any LLM to reason, decide and act on open-ended tasks. Every agent comes with memory, tool calling, MCP, logging, tracing and eval primitives built in. Build and test agents in a local playground, then deploy to production with the same configuration.

You are an enterprise search agent. You should search across internal wikis, Slack threads, and PDFs. Prioritize context based on the role of the user (such as marketing, sales, etc.)

You are a copywriting agent. Based on the current marketing copy and website, write additional content and generate an image given the product specifications of the user.

Built-in agent primitives

When you need agents that retain context, call tools and evaluate their own output, Mastra gives you every primitive built in. Tool calling connects agents to external APIs and services. Memory retains conversation history, user data and relevant context across interactions. Logging, tracing and evals surface what happened and why.

Tool calls and MCP

Orchestrate agent networks

Single agents handle open-ended tasks well. Mastra lets you compose agents with workflows for increased reliability, or assemble agent networks where agents collaborate on more complex problems. Specify execution paths for increased control or delegate to a team of agents for increased power.

Agentic workflows

Specify execution paths across agents and tool calls for increased control

docs

Agent networks

Assemble a team of agents to tackle complex problems together

docs

Context and guardrails

Mastra gives you precise control over what agents know and what they can do. Dynamically inject context based on the request, configure access from user identity and enforce guardrails against prompt injection and data leaks. Suspend and resume agent execution for human review.

Human in the loop

Python trains,
TypeScript ships.

npm create mastra

Contact Sales

Stateful AI agents with any LLM.

Built-in agent primitives

Execute real-world actions through external APIs and services

Recall relevant context across conversations and store persistent user data

Trace agent execution and score output quality across every run

Orchestrate agent networks

Specify execution paths across agents and tool calls for increased control

Assemble a team of agents to tackle complex problems together

Context and guardrails

Suspend and resume agent execution for human review at any point

Configure agent access and behavior dynamically from user context

Guard agents against prompt injection and prevent data leaks

Python trains,
TypeScript ships.

Stateful AI agents with any LLM.

Built-in agent primitives

Execute real-world actions through external APIs and services

Recall relevant context across conversations and store persistent user data

Trace agent execution and score output quality across every run

Orchestrate agent networks

Specify execution paths across agents and tool calls for increased control

Assemble a team of agents to tackle complex problems together

Context and guardrails

Suspend and resume agent execution for human review at any point

Configure agent access and behavior dynamically from user context

Guard agents against prompt injection and prevent data leaks

Python trains, TypeScript ships.

Python trains,
TypeScript ships.