# Celesto AI

> Celesto provides secure, isolated computers that AI agents can use to browse, code, and get real work done.

## Docs

- [Agent-to-Agent communication](https://docs.celesto.ai/agentor/agent-to-agent.md)
- [Agents API](https://docs.celesto.ai/agentor/agents-api.md): Build an Agent, add capabilities and serve as an API in just few lines of code.
- [A2AController](https://docs.celesto.ai/agentor/api/a2a/controller.md): A2A protocol controller for agent-to-agent communication
- [A2A Types](https://docs.celesto.ai/agentor/api/a2a/types.md): Type definitions for the A2A protocol
- [Agentor](https://docs.celesto.ai/agentor/api/agentor.md): Core class for building and deploying production-ready AI agents
- [LLM](https://docs.celesto.ai/agentor/api/llm.md): Lightweight LLM client for direct model interactions
- [CelestoMCPHub](https://docs.celesto.ai/agentor/api/mcp/hub.md): Client for connecting to Celesto AI's MCP Hub service
- [LiteMCP](https://docs.celesto.ai/agentor/api/mcp/litemcp.md): ASGI-compatible MCP server built on FastAPI with decorator-based API
- [MCPAPIRouter](https://docs.celesto.ai/agentor/api/mcp/router.md): Router for MCP JSON-RPC methods with FastAPI-like decorator API
- [ModelSettings](https://docs.celesto.ai/agentor/api/model-settings.md): Configure LLM model parameters and behavior
- [Tools](https://docs.celesto.ai/agentor/api/tools.md): Create and use tools with AI agents
- [Calculator Tool](https://docs.celesto.ai/agentor/api/tools/calculator.md): Perform basic arithmetic operations
- [FetchTool](https://docs.celesto.ai/agentor/api/tools/fetch.md): Fetch content from URLs using HTTP methods
- [GitTool](https://docs.celesto.ai/agentor/api/tools/git.md): Execute Git commands on repositories
- [GitHubTool](https://docs.celesto.ai/agentor/api/tools/github.md): Interact with GitHub repositories, issues, and pull requests
- [GmailTool](https://docs.celesto.ai/agentor/api/tools/gmail.md): Read-only Gmail tool for searching and fetching messages
- [CalendarTool](https://docs.celesto.ai/agentor/api/tools/google-calendar.md): Manage Google Calendar events, find free slots, and invite guests
- [Shell Tool](https://docs.celesto.ai/agentor/api/tools/shell.md): Execute shell commands securely from agents
- [Weather Tool](https://docs.celesto.ai/agentor/api/tools/weather.md): Get current weather information for any location
- [Web Search Tool](https://docs.celesto.ai/agentor/api/tools/web-search.md): Search the web using OpenAI's built-in web search capability
- [Agent-to-Agent (A2A) Protocol](https://docs.celesto.ai/agentor/concepts/a2a-protocol.md): Enable standardized communication between AI agents with the A2A protocol
- [Agent Architecture](https://docs.celesto.ai/agentor/concepts/agents.md): Understand the Agentor agent lifecycle, configuration, and execution model
- [Deployment](https://docs.celesto.ai/agentor/concepts/deployment.md): Deploy Agentor agents and MCP servers to production with Celesto AI or self-hosted infrastructure
- [Model Context Protocol (MCP)](https://docs.celesto.ai/agentor/concepts/mcp.md): Build and integrate MCP servers with LiteMCP - a FastAPI-native implementation
- [Skills System](https://docs.celesto.ai/agentor/concepts/skills.md): Dynamically load specialized instructions and resources to improve agent performance on specific tasks
- [Tool System](https://docs.celesto.ai/agentor/concepts/tools.md): Build and integrate tools for agent capabilities using the extensible tool registry
- [Run agents in cloud](https://docs.celesto.ai/agentor/deploy.md)
- [Agent-to-Agent Communication](https://docs.celesto.ai/agentor/guides/agent-communication.md): Enable agents to communicate and collaborate using the A2A protocol
- [Building Agents](https://docs.celesto.ai/agentor/guides/building-agents.md): Learn how to build production-ready AI agents with Agentor
- [Creating Custom Tools](https://docs.celesto.ai/agentor/guides/custom-tools.md): Build and integrate custom tools to extend agent capabilities
- [Deploying with Celesto CLI](https://docs.celesto.ai/agentor/guides/deployment.md): Deploy production-ready agents to the cloud with a single command
- [Building MCP Servers](https://docs.celesto.ai/agentor/guides/mcp-servers.md): Create Model Context Protocol servers with LiteMCP for tool sharing and integration
- [Tracing and Monitoring](https://docs.celesto.ai/agentor/guides/observability.md): Monitor agent performance, debug issues, and optimize costs with built-in observability
- [Streaming Responses](https://docs.celesto.ai/agentor/guides/streaming.md): Stream agent responses in real-time for better user experience
- [Agentor](https://docs.celesto.ai/agentor/home.md): Build, deploy, and operate production-ready AI agents with tool use, MCP support, and agent-to-agent communication.
- [Installation](https://docs.celesto.ai/agentor/installation.md): Install Agentor and configure your development environment
- [Introduction to Agentor](https://docs.celesto.ai/agentor/introduction.md): Build and deploy production-ready AI agents with MCP server support, A2A protocol, and serverless deployment
- [Quickstart](https://docs.celesto.ai/agentor/quickstart.md): Build your first AI agent with Agentor in under 5 minutes
- [Agent Skills](https://docs.celesto.ai/agentor/skills.md): Skills are directories containing instructions, scripts, and resources that an agent loads dynamically to improve performance on specialized tasks.
- [Build a custom MCP Server](https://docs.celesto.ai/agentor/tools/LiteMCP.md): Agents can access internal data sources and APIs with a custom MCP Server.
- [LiteMCP and FastAPI APIRouter](https://docs.celesto.ai/agentor/tools/LiteMCP-fastapi.md): Build MCP Servers with a FastAPI APIRouter.
- [Authorization in MCP](https://docs.celesto.ai/agentor/tools/auth.md): Learn about authorization and security in MCP
- [Tool-use and MCP Server](https://docs.celesto.ai/agentor/tools/overview.md): Learn about how LLMs make use of tools and MCP Servers.
- [Tool Search](https://docs.celesto.ai/agentor/tools/tool-search.md): Reduce LLM tool context bloat by dynamically filtering LLM tools using Tool Search API.
- [Tool use with Agentor](https://docs.celesto.ai/agentor/tools/tool-use.md): Learn about tool use with Agentor.
- [Tracing and Observability](https://docs.celesto.ai/agentor/tracing.md): Capture and inspect agent runs and tool calls with Celesto tracing.
- [What is an SDR Agent?](https://docs.celesto.ai/enterprise/use-cases/SDR-Agent.md): An SDR Agent (Sales Development Representative Agent) is an AI-powered assistant that automates lead research, outreach, qualification, and scheduling. It identifies target prospects, crafts personalized messages, follows up with context awareness, and updates your CRM automatically. The agent opera…
- [What is a Customer Support Agent?](https://docs.celesto.ai/enterprise/use-cases/customer-support-Agent.md): A Customer Support Agent is an AI-powered assistant that manages customer queries across chat, email, and support portals. It understands intent, fetches information from your knowledge base, executes actions such as order tracking or refunds, and responds instantly across channels.
- [What is an E-commerce Agent?](https://docs.celesto.ai/enterprise/use-cases/ecommerce-Agent.md): An E-commerce Agent is an AI-powered assistant that manages product discovery, order tracking, returns, recommendations, and upselling through natural conversation. It connects to your commerce systems, understands customer intent, and delivers personalized responses across chat, voice, or in-app in…
- [What is an HR & People Ops Agent?](https://docs.celesto.ai/enterprise/use-cases/human-resource-Agent.md): An HR & People Ops Agent is an AI-powered assistant for internal teams that automates repetitive HR and People Operations workflows. It answers policy questions, assists with onboarding, manages internal requests, and connects directly with HR tools and documentation. It functions as the conversatio…
- [Technical architecture](https://docs.celesto.ai/smolvm/advanced/architecture.md): Deep dive into SmolVM's architecture, components, and design decisions
- [Performance benchmarks](https://docs.celesto.ai/smolvm/advanced/performance.md): Performance characteristics, optimization tips, and benchmark results for SmolVM
- [Troubleshooting guide](https://docs.celesto.ai/smolvm/advanced/troubleshooting.md): Common issues, error messages, and solutions for SmolVM
- [BrowserSession](https://docs.celesto.ai/smolvm/api/browsersession.md): Launch and control an isolated browser inside a sandbox
- [BrowserSessionConfig](https://docs.celesto.ai/smolvm/api/browsersessionconfig.md): Configure browser session settings including viewport, recording, and resources
- [BrowserSessionInfo](https://docs.celesto.ai/smolvm/api/browsersessioninfo.md): Runtime information for an active browser session
- [CommandResult](https://docs.celesto.ai/smolvm/api/commandresult.md): Result object returned from executing commands on a guest VM via SSH
- [Exceptions](https://docs.celesto.ai/smolvm/api/exceptions.md): Exception classes for error handling in SmolVM
- [HostManager](https://docs.celesto.ai/smolvm/api/hostmanager.md): Validates host environment and manages the Firecracker binary for SmolVM.
- [ImageBuilder](https://docs.celesto.ai/smolvm/api/imagebuilder.md): Build custom VM images with SSH pre-configured using Docker
- [ImageManager](https://docs.celesto.ai/smolvm/api/imagemanager.md): Download, cache, and verify pre-built VM images
- [ImageSource](https://docs.celesto.ai/smolvm/api/imagesource.md): Define downloadable VM image metadata with verification
- [InternetSettings](https://docs.celesto.ai/smolvm/api/internetsettings.md): Control which domains a sandbox can access
- [NetworkConfig](https://docs.celesto.ai/smolvm/api/networkconfig.md): Network configuration model for SmolVM guests.
- [SmolVM](https://docs.celesto.ai/smolvm/api/smolvm.md): High-level interface for managing a single microVM with automatic configuration and lifecycle management
- [SnapshotInfo](https://docs.celesto.ai/smolvm/api/snapshotinfo.md): Metadata for a saved VM snapshot, including file paths and restore status
- [SSHClient](https://docs.celesto.ai/smolvm/api/sshclient.md): Execute commands on microVM guests via persistent SSH connections.
- [VMConfig](https://docs.celesto.ai/smolvm/api/vmconfig.md): Configuration model for creating a microVM with resource allocation, disk settings, and boot parameters
- [VMInfo](https://docs.celesto.ai/smolvm/api/vminfo.md): Runtime information about a VM including status, configuration, network details, and process information
- [smolvm cleanup](https://docs.celesto.ai/smolvm/cli/cleanup.md): Clean stale SmolVM resources and VMs
- [smolvm create](https://docs.celesto.ai/smolvm/cli/create.md): Create an SSH-ready VM and leave it running
- [smolvm doctor](https://docs.celesto.ai/smolvm/cli/doctor.md): Run host diagnostics for the selected backend
- [smolvm env](https://docs.celesto.ai/smolvm/cli/env.md): Manage environment variables on a running VM
- [smolvm list](https://docs.celesto.ai/smolvm/cli/list.md): List your SmolVM instances and check their current status
- [CLI overview](https://docs.celesto.ai/smolvm/cli/overview.md): Overview of the SmolVM command-line interface for managing sandboxes and browser sessions
- [smolvm snapshot](https://docs.celesto.ai/smolvm/cli/snapshot.md): Create, restore, list, and delete Firecracker VM snapshots from the command line
- [smolvm ssh](https://docs.celesto.ai/smolvm/cli/ssh.md): SSH into an existing VM, auto-starting it if needed
- [smolvm stop](https://docs.celesto.ai/smolvm/cli/stop.md): Stop a running VM and release its resources
- [smolvm ui](https://docs.celesto.ai/smolvm/cli/ui.md): Start the SmolVM dashboard UI server
- [Firecracker vs QEMU backends](https://docs.celesto.ai/smolvm/concepts/backends.md): Choose the right microVM backend for your platform and understand their trade-offs
- [Network configuration](https://docs.celesto.ai/smolvm/concepts/networking.md): How SmolVM connects sandboxes to the internet while keeping them isolated from each other
- [Architecture overview](https://docs.celesto.ai/smolvm/concepts/overview.md): How SmolVM keeps your code isolated and your host safe
- [Security model](https://docs.celesto.ai/smolvm/concepts/security.md): Understand what SmolVM protects, what it does not, and how to run it safely
- [AI agent integration](https://docs.celesto.ai/smolvm/guides/ai-agent-integration.md): Use SmolVM as a secure sandbox tool for PydanticAI, OpenAI Agents, LangChain, and other AI frameworks
- [Basic usage](https://docs.celesto.ai/smolvm/guides/basic-usage.md): Create sandboxes, run commands, and check results with SmolVM
- [Custom images](https://docs.celesto.ai/smolvm/guides/custom-images.md): Build custom VM images with your own tools and dependencies pre-installed
- [Environment variables](https://docs.celesto.ai/smolvm/guides/environment-variables.md): Pass configuration and secrets into SmolVM sandboxes
- [Port forwarding](https://docs.celesto.ai/smolvm/guides/port-forwarding.md): Access web servers, databases, and other services running inside a sandbox from your host
- [Snapshots](https://docs.celesto.ai/smolvm/guides/snapshots.md): Save and restore the full state of a running sandbox so you can resume it later
- [VM lifecycle management](https://docs.celesto.ai/smolvm/guides/vm-lifecycle.md): Create, start, stop, pause, and reconnect to SmolVM sandboxes
- [Installation](https://docs.celesto.ai/smolvm/installation.md): Install SmolVM on Linux or macOS
- [Introduction](https://docs.celesto.ai/smolvm/introduction.md): Run code, start a browser, and give AI agents an isolated workspace with SmolVM
- [Quickstart](https://docs.celesto.ai/smolvm/quickstart.md): Set up SmolVM and run your first sandbox in under five minutes

## Optional

- [Get started](https://celesto.ai)
- [GitHub](https://github.com/celestoai)