New Agent Capabilities: Code Execution and RAG Search

Our agents can now execute code in isolated sandboxes and perform semantic search across your documents.

Code Execution Now Live

Agents can run Python, JavaScript, and TypeScript code in secure, ephemeral sandboxes. This enables data analysis, API integrations, and custom logic execution.

Retrieval Augmented Generation gives agents access to your documents. Upload PDFs, DOCX, or text files and your agent can search them semantically.

How It Works

Enable these tools in your agent’s configuration panel. No setup required — everything runs server-side.

Back to news