# 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 with the A2A Protocol](https://docs.celesto.ai/agentor/agent-to-agent.md): Use the Agent-to-Agent (A2A) protocol to let AI agents discover each other, share agent cards, and exchange JSON-RPC messages across frameworks.
- [Agents API](https://docs.celesto.ai/agentor/agents-api.md): Build an Agentor agent in Python, attach tools and skills, and serve it as a production-ready REST API with just a few lines of code.
- [A2AController](https://docs.celesto.ai/agentor/api/a2a/controller.md): A2AController class reference: a FastAPI router that exposes an agent card manifest and handles JSON-RPC messaging for the A2A protocol v0.3.0.
- [A2A Types](https://docs.celesto.ai/agentor/api/a2a/types.md): Reference for A2A protocol type definitions including JSONRPCRequest, JSONRPCResponse, AgentCard, AgentSkill, and message types from the a2a-sdk.
- [Agentor class API reference](https://docs.celesto.ai/agentor/api/agentor.md): Agentor class API reference: configure model, instructions, tools, skills, and tracing, then run agents synchronously, asynchronously, or as a streamed API.
- [LLM](https://docs.celesto.ai/agentor/api/llm.md): LLM class reference: a lightweight client for calling OpenAI, Anthropic, Gemini, and other LiteLLM models directly without the full agent framework.
- [CelestoMCPHub](https://docs.celesto.ai/agentor/api/mcp/hub.md): CelestoMCPHub reference: a context-managed client that aggregates multiple MCP servers and exposes their tools to your agents through a single endpoint.
- [LiteMCP](https://docs.celesto.ai/agentor/api/mcp/litemcp.md): LiteMCP class reference: build ASGI-compatible Model Context Protocol servers on FastAPI using decorators for tools, prompts, and resources.
- [MCPAPIRouter](https://docs.celesto.ai/agentor/api/mcp/router.md): MCPAPIRouter reference: register MCP tools, prompts, and resources with FastAPI-style decorators and automatic JSON-RPC schema generation.
- [ModelSettings](https://docs.celesto.ai/agentor/api/model-settings.md): ModelSettings reference: tune temperature, max tokens, top-p, and other LLM generation parameters when configuring an Agentor agent.
- [Tools](https://docs.celesto.ai/agentor/api/tools.md): Reference for the Agentor tools system: function decorators, BaseTool classes, and global tool registry to extend agent capabilities.
- [Calculator Tool](https://docs.celesto.ai/agentor/api/tools/calculator.md): CalculatorTool reference: give Agentor agents add, subtract, multiply, and divide capabilities for precise numerical calculations during reasoning tasks.
- [FetchTool](https://docs.celesto.ai/agentor/api/tools/fetch.md): FetchTool reference: let agents make HTTP GET, POST, PUT, and DELETE requests to retrieve web pages, call APIs, and fetch URL content.
- [GitTool](https://docs.celesto.ai/agentor/api/tools/git.md): GitTool API reference: let agents clone repositories, commit and push changes, manage branches, and run Git commands on local working trees.
- [GitHubTool](https://docs.celesto.ai/agentor/api/tools/github.md): GitHubTool reference: agents can read repositories, create issues, open pull requests, and call the GitHub API using PyGithub and an access token.
- [GmailTool](https://docs.celesto.ai/agentor/api/tools/gmail.md): GmailTool reference: give agents read-only Gmail access to list, search, and fetch email messages using Google OAuth2 and the Gmail API.
- [CalendarTool](https://docs.celesto.ai/agentor/api/tools/google-calendar.md): CalendarTool reference: agents can read, create, and delete Google Calendar events, find free time slots, and invite guests for scheduling workflows.
- [ScrapeGraphAI](https://docs.celesto.ai/agentor/api/tools/scrapegraphai.md): ScrapeGraphAI tool reference: agents scrape pages, extract structured data, run web searches, crawl sites, and schedule change monitors via the scrapegraph-py 2.x SDK.
- [Shell Tool](https://docs.celesto.ai/agentor/api/tools/shell.md): ShellTool reference: let agents run bash and shell commands inside a controlled execution environment with timeouts and a custom executor callback.
- [Weather Tool](https://docs.celesto.ai/agentor/api/tools/weather.md): GetWeatherTool reference: agents fetch real-time weather data for any city or location using the WeatherAPI.com service and a free API key.
- [Web Search Tool](https://docs.celesto.ai/agentor/api/tools/web-search.md): WebSearchTool reference: add real-time web search to your agents using OpenAI's built-in search capability — no extra API keys required.
- [Agent-to-Agent (A2A) Protocol](https://docs.celesto.ai/agentor/concepts/a2a-protocol.md): How the Agent-to-Agent (A2A) protocol enables standardized JSON-RPC messaging, agent card discovery, and interoperability between AI agent frameworks.
- [Agent Architecture](https://docs.celesto.ai/agentor/concepts/agents.md): Agent architecture in Agentor: how the Agent class handles configuration, tool registration, model calls, lifecycle, and synchronous or streaming execution.
- [Deployment](https://docs.celesto.ai/agentor/concepts/deployment.md): Deploy Agentor agents and MCP servers to production using Celesto AI, self-hosted uvicorn or gunicorn servers, Docker containers, or serverless platforms.
- [Model Context Protocol (MCP)](https://docs.celesto.ai/agentor/concepts/mcp.md): Learn how the Model Context Protocol (MCP) connects LLMs to external tools, prompts, and resources, and how LiteMCP implements MCP natively on FastAPI.
- [Skills System](https://docs.celesto.ai/agentor/concepts/skills.md): How Agentor's skills system dynamically loads specialized instructions, scripts, and resources so agents perform better on focused, domain-specific tasks.
- [Tool System](https://docs.celesto.ai/agentor/concepts/tools.md): How the Agentor tool system works: function tools, BaseTool classes, the global registry, and integration with external MCP servers for agent capabilities.
- [Run agents in cloud](https://docs.celesto.ai/agentor/deploy.md): Serve Agentor agents as production REST APIs and deploy them to Celesto's cloud or your own infrastructure with the built-in serve command.
- [Agent-to-Agent Communication](https://docs.celesto.ai/agentor/guides/agent-communication.md): Step-by-step guide to enabling agent-to-agent collaboration with the A2A protocol — agent cards, JSON-RPC messaging, streaming, and multi-agent workflows.
- [Building Agents](https://docs.celesto.ai/agentor/guides/building-agents.md): Build production-ready AI agents with Agentor: configure models, attach tools, add instructions, handle streaming responses, and serve them as APIs.
- [Creating Custom Tools](https://docs.celesto.ai/agentor/guides/custom-tools.md): Build custom tools for Agentor agents using the function_tool decorator or BaseTool class to integrate APIs, databases, and external services.
- [Deploying with Celesto CLI](https://docs.celesto.ai/agentor/guides/deployment.md): Deploy Agentor agents and MCP servers to the cloud using the Celesto CLI, with built-in observability, automatic scaling, and a single deploy command.
- [Building MCP Servers](https://docs.celesto.ai/agentor/guides/mcp-servers.md): Build Model Context Protocol (MCP) servers with LiteMCP and FastAPI to share tools, prompts, and resources across agents and MCP-compatible clients.
- [Tracing and Monitoring](https://docs.celesto.ai/agentor/guides/observability.md): Monitor Agentor agents in production with built-in tracing — debug tool calls, track token usage and costs, and analyze multi-agent workflows.
- [Streaming Responses](https://docs.celesto.ai/agentor/guides/streaming.md): Stream Agentor agent responses token by token using stream_chat to lower perceived latency, show tool calls live, and support cancellation in your UI.
- [Agentor: Python framework for long-running AI agents](https://docs.celesto.ai/agentor/home.md): Agentor is an open-source Python framework for building, deploying, and operating long-running AI agents with tool use, MCP, and the A2A protocol.
- [Install Agentor and configure LLM providers](https://docs.celesto.ai/agentor/installation.md): Install the Agentor Python framework on macOS, Linux, or Windows using pip or uv, and configure LLM provider API keys for OpenAI, Anthropic, or Gemini.
- [Build your first AI agent with Agentor](https://docs.celesto.ai/agentor/quickstart.md): Build your first AI agent with Agentor in under five minutes — install the framework, write a weather agent, attach tools, and run your first query.
- [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): Build a custom MCP server with LiteMCP so agents can access your internal data sources and APIs through the Model Context Protocol over FastAPI.
- [LiteMCP and FastAPI APIRouter](https://docs.celesto.ai/agentor/tools/LiteMCP-fastapi.md): Add Model Context Protocol endpoints to an existing FastAPI app with LiteMCP's MCPAPIRouter, including CORS setup and tool registration patterns.
- [Authorization in MCP](https://docs.celesto.ai/agentor/tools/auth.md): Authorization patterns for MCP servers — host-authenticated tools, single-tenant deployments, multi-tenant access, and securing client credentials.
- [Tool-use and MCP Server](https://docs.celesto.ai/agentor/tools/overview.md): Introduction to LLM tool use and MCP servers — how language models call external APIs, choose tools, and use the Model Context Protocol for integrations.
- [Tool Search](https://docs.celesto.ai/agentor/tools/tool-search.md): Cut LLM tool context bloat by dynamically filtering relevant tools at runtime using Agentor's Tool Search API — works with any LLM provider.
- [Tool use with Agentor](https://docs.celesto.ai/agentor/tools/tool-use.md): Use the @function_tool decorator in Agentor to give an agent access to external APIs — schema is generated from your Python function automatically.
- [Tracing and Observability](https://docs.celesto.ai/agentor/tracing.md): Capture and inspect agent runs, tool calls, and multi-step traces in the Celesto dashboard to debug failures and monitor agent behavior in production.
- [Authenticate the Celesto SDK and CLI](https://docs.celesto.ai/celesto-sdk/authentication.md): Authenticate Celesto SDK and CLI calls with the CELESTO_API_KEY environment variable, direct credentials, or the saved CLI login session.
- [Celesto CLI reference](https://docs.celesto.ai/celesto-sdk/cli.md): Use the celesto CLI to authenticate, create computers, run commands, publish ports, and manage sandbox lifecycle from the terminal.
- [Create and manage sandboxed computers](https://docs.celesto.ai/celesto-sdk/computers.md): Use the Celesto Computers API to create sandboxed computers, select templates, run shell commands, publish ports, and manage the VM lifecycle.
- [Deploy agents on Celesto](https://docs.celesto.ai/celesto-sdk/deployments.md): Learn the current status of Celesto deployment APIs and use sandboxed computers while the deployment resource is updated.
- [Handle Celesto SDK errors](https://docs.celesto.ai/celesto-sdk/errors.md): Catch and handle Celesto SDK authentication, validation, not found, rate limit, server, and network exceptions across the Python and JavaScript clients.
- [Start, stop, and delete computers](https://docs.celesto.ai/celesto-sdk/features/lifecycle.md): Manage Celesto computer lifecycle with the SDK: create, stop, start, and delete workspaces so you can pause tasks, resume later, and free resources.
- [Keep work between sessions](https://docs.celesto.ai/celesto-sdk/features/persistence.md): Persist files, installed packages, and workspace changes across sessions by stopping a Celesto computer and restarting it later by ID or name.
- [Petabyte-scale storage for AI agent sandboxes](https://docs.celesto.ai/celesto-sdk/features/petabyte-storage.md): Use CelestoFS to give AI agent sandboxes a durable petabyte-scale filesystem for large files, repositories, build artifacts, and long-running workspace state.
- [Publish ports from a sandboxed computer](https://docs.celesto.ai/celesto-sdk/features/publish-ports.md): Expose a web server, API, notebook, or preview app running inside a Celesto computer through a public HTTPS URL with the SDK or CLI.
- [Choose computer size](https://docs.celesto.ai/celesto-sdk/features/resources.md): Configure CPU, memory, disk size, and templates when creating a Celesto computer so coding agents have enough resources for builds, packages, and apps.
- [Connect users to third-party OAuth with Gatekeeper](https://docs.celesto.ai/celesto-sdk/gatekeeper.md): Use Celesto Gatekeeper to connect your end users to third-party providers like Google Drive with delegated OAuth access scoped to your project.
- [Run Pi coding agent in the cloud](https://docs.celesto.ai/celesto-sdk/guides/pi-coding-agent.md): Run Pi coding agent remotely on a Celesto cloud computer, keep model credentials local, push and sync project files explicitly, and troubleshoot common setup issues.
- [Run a GitHub app on a Celesto computer](https://docs.celesto.ai/celesto-sdk/guides/publish-github-app.md): Clone a GitHub web app onto a Celesto computer, install dependencies, run it on port 8000, and share a public HTTPS preview URL.
- [Sandbox an OpenAI agent with Celesto or SmolVM](https://docs.celesto.ai/celesto-sdk/openai-agents.md): Give an OpenAI agent its own sandboxed computer. Use a hosted Celesto computer or a local SmolVM as the workspace for a SandboxAgent.
- [Celesto SDK for sandboxed computers and agents](https://docs.celesto.ai/celesto-sdk/overview.md): Use the Celesto SDK to create sandboxed computers, run shell commands, publish ports, and deploy AI agents from either Python or JavaScript clients.
- [Get started with the Celesto SDK](https://docs.celesto.ai/celesto-sdk/quickstart.md): Install the Celesto SDK, authenticate with an API key, create your first sandboxed computer, run a command in it, and clean up the resources.
- [Agent computers](https://docs.celesto.ai/getting-started/agent-computers.md): Understand Celesto agent computers: isolated sandboxes where AI agents can run commands, read and write files, publish ports, and keep state.
- [Browser agents](https://docs.celesto.ai/getting-started/browser-agents.md): Use Celesto and SmolVM browser sandboxes for AI agents that browse sites, test web apps, inspect pages, and run computer-use workflows.
- [Coding agents](https://docs.celesto.ai/getting-started/coding-agents.md): Build coding agents on Celesto with sandboxed terminals, repositories, package installs, tests, preview apps, durable state, and large workspace storage.
- [Durable workspaces](https://docs.celesto.ai/getting-started/durable-workspaces.md): Learn how Celesto keeps agent state durable across sessions with root disk persistence and petabyte-scale CelestoFS workspace storage.
- [Hosted and open source](https://docs.celesto.ai/getting-started/hosted-and-open-source.md): Choose between the hosted Celesto Platform and Celesto's open-source projects: SmolVM, SmolFS, and Agentor.
- [Long-running agents](https://docs.celesto.ai/getting-started/long-running-agents.md): Design long-running AI agent jobs with Celesto durable computers, stop/start lifecycle, saved workspace state, and large storage for artifacts.
- [Why Celesto](https://docs.celesto.ai/getting-started/why-celesto.md): Learn why Celesto gives AI agents secure sandboxed computers with durable state, petabyte-scale workspace storage, browser access, ports, and open-source runtimes.
- [Computer for AI Agents](https://docs.celesto.ai/index.md): Create secure sandboxed computers for AI agents with persistent files, CelestoFS storage, browser access, terminals, ports, SDKs, and open-source runtimes.
- [SmolFS CLI](https://docs.celesto.ai/smolfs/cli.md): Use the SmolFS command line to check your machine, create volumes, open workspace folders, save writes, and close them.
- [Install SmolFS](https://docs.celesto.ai/smolfs/installation.md): Install the SmolFS CLI, check local mount support, and add Python or TypeScript bindings for agent code.
- [SmolFS: durable workspace folders for AI agents](https://docs.celesto.ai/smolfs/overview.md): Learn what SmolFS does, when to use it, and where to start with the first open-source release.
- [Create your first SmolFS workspace](https://docs.celesto.ai/smolfs/quickstart.md): Create a local SmolFS workspace, write a file, close it, and open it again.
- [SmolFS SDKs](https://docs.celesto.ai/smolfs/sdk.md): Use SmolFS from Python or TypeScript when an agent runner should manage durable workspace folders from code.
- [Performance benchmarks](https://docs.celesto.ai/smolvm/advanced/performance.md): SmolVM benchmark numbers for boot time, control-channel readiness, and command latency, plus tuning tips for fast agent workloads.
- [Troubleshooting guide](https://docs.celesto.ai/smolvm/advanced/troubleshooting.md): Troubleshoot common SmolVM issues — KVM permissions, networking failures, image download errors, SSH timeouts — using smolvm doctor and verified fixes.
- [BootImage and boot helpers](https://docs.celesto.ai/smolvm/api/bootimage.md): BootImage, DirectKernelBoot, and FirmwareBoot reference for describing custom SmolVM disk images and the kernel or firmware boot path the VM should use.
- [Browser and desktop sandboxes](https://docs.celesto.ai/smolvm/api/browsersession.md): Start browser and desktop sandboxes with SmolVM.browser() and SmolVM.desktop(), then connect automation tools or open a live viewer.
- [Browser and desktop options](https://docs.celesto.ai/smolvm/api/browsersessionconfig.md): Options for SmolVM.browser() and SmolVM.desktop(), including viewport size, profiles, recording, resources, and shared folders.
- [Display sandbox object](https://docs.celesto.ai/smolvm/api/browsersessioninfo.md): Reference for the object returned by SmolVM.browser() and SmolVM.desktop(), including URLs, IDs, lifecycle methods, and browser helpers.
- [Callback hooks Python reference](https://docs.celesto.ai/smolvm/api/callbacks.md): Reference for SmolVM's Callback base class, RunContext dataclass, and CommandBlockedError — the pre-run, post-run, and error hooks that fire around run().
- [CommandResult](https://docs.celesto.ai/smolvm/api/commandresult.md): CommandResult reference: an immutable Pydantic model with exit_code, stdout, stderr, and helper properties returned by SmolVM SSH command execution.
- [DockerRootfsBuilder](https://docs.celesto.ai/smolvm/api/dockerrootfsbuilder.md): DockerRootfsBuilder reference: build and cache raw ext4 root filesystems from Dockerfiles and return BootImage metadata for SmolVM.from_image.
- [Exceptions](https://docs.celesto.ai/smolvm/api/exceptions.md): Reference for SmolVM exception classes — SmolVMError, ValidationError, VMNotFoundError, NetworkError, FirecrackerAPIError — for robust error handling.
- [HostManager](https://docs.celesto.ai/smolvm/api/hostmanager.md): HostManager reference: validate KVM access and host networking tools, install and pin the Firecracker binary, and prepare the SmolVM runtime environment.
- [ImageBuilder](https://docs.celesto.ai/smolvm/api/imagebuilder.md): ImageBuilder reference: build minimal Alpine or Debian-based VM rootfs images with SSH pre-configured using Docker, complete with networking and init scripts.
- [ImageManager](https://docs.celesto.ai/smolvm/api/imagemanager.md): ImageManager reference: download pre-built SmolVM kernel and rootfs images, cache them locally, and verify integrity with SHA-256 checksums.
- [ImageSource](https://docs.celesto.ai/smolvm/api/imagesource.md): ImageSource reference: define metadata for downloadable SmolVM kernel and rootfs images, including URLs, file names, and SHA-256 verification hashes.
- [InternetSettings](https://docs.celesto.ai/smolvm/api/internetsettings.md): InternetSettings reference: restrict outbound network access from a SmolVM sandbox to an allowlist of domains using the VMConfig internet_settings field.
- [Kernel helpers](https://docs.celesto.ai/smolvm/api/kernelhelpers.md): ensure_base_kernel_for_backend reference for resolving, downloading, and caching the SmolVM base kernel matching a chosen backend and CPU architecture.
- [NetworkConfig](https://docs.celesto.ai/smolvm/api/networkconfig.md): NetworkConfig reference: an immutable Pydantic model defining a guest VM's IP, gateway, TAP device, MAC address, and SSH port forwarding settings.
- [SmolVM Python class reference](https://docs.celesto.ai/smolvm/api/smolvm.md): SmolVM class reference: the Python API for creating microVM sandboxes, running commands over the selected control channel, and managing the full sandbox lifecycle.
- [SnapshotInfo](https://docs.celesto.ai/smolvm/api/snapshotinfo.md): SnapshotInfo reference: read snapshot metadata including ID, backend, artifacts, source VM config, creation time, type, and restore status.
- [SSHClient](https://docs.celesto.ai/smolvm/api/sshclient.md): SSHClient reference: run commands on SmolVM guest VMs through persistent paramiko SSH connections, eliminating the per-call ssh process overhead.
- [VMConfig Python reference](https://docs.celesto.ai/smolvm/api/vmconfig.md): VMConfig reference: the Pydantic model that defines CPU, memory, kernel, rootfs, disk, network, and boot parameters for a SmolVM microVM.
- [VMInfo](https://docs.celesto.ai/smolvm/api/vminfo.md): VMInfo reference: a Pydantic model exposing a SmolVM's lifecycle state, configuration, networking details, process IDs, and runtime metadata.
- [smolvm sandbox delete](https://docs.celesto.ai/smolvm/cli/cleanup.md): Use smolvm sandbox delete to remove one sandbox or clean up every sandbox on the host with a confirmation step.
- [smolvm sandbox create](https://docs.celesto.ai/smolvm/cli/create.md): Use smolvm sandbox create to start a new sandbox from the command line, name it, choose an operating system or image, share folders, and get JSON output for scripts.
- [smolvm doctor](https://docs.celesto.ai/smolvm/cli/doctor.md): Run smolvm doctor to verify your host meets all SmolVM prerequisites — KVM access, required binaries, network tools, and the chosen Firecracker or QEMU backend.
- [smolvm sandbox env](https://docs.celesto.ai/smolvm/cli/env.md): Use smolvm sandbox env to set, remove, and list environment variables inside a running sandbox over SSH or the SmolVM guest-agent channel.
- [smolvm sandbox file](https://docs.celesto.ai/smolvm/cli/file.md): Use smolvm sandbox file to upload files into a sandbox or download files back to your host over SSH or the SmolVM guest-agent channel.
- [smolvm sandbox list](https://docs.celesto.ai/smolvm/cli/list.md): Use smolvm sandbox list to view running, stopped, paused, and errored sandboxes, filter by lifecycle state, and produce JSON for scripts.
- [CLI overview](https://docs.celesto.ai/smolvm/cli/overview.md): Overview of the SmolVM command-line interface for creating sandboxes, opening shells, managing snapshots, starting the HTTP server, and launching browser or desktop sessions.
- [smolvm sandbox port](https://docs.celesto.ai/smolvm/cli/port.md): Use smolvm sandbox port to share a service running inside a sandbox on localhost, list active forwards, and close them when finished.
- [smolvm prune](https://docs.celesto.ai/smolvm/cli/prune.md): Reclaim disk space by deleting cached SmolVM images from older releases. Preview targets with --dry-run or get machine-readable output with --json.
- [smolvm server](https://docs.celesto.ai/smolvm/cli/server.md): Use smolvm server start to run the local SmolVM HTTP API for scripts, service integrations, and the TypeScript SDK.
- [smolvm sandbox shell](https://docs.celesto.ai/smolvm/cli/shell.md): Use smolvm sandbox shell to open the fastest interactive shell SmolVM can provide, with guest-agent terminal access on recent images and automatic start or resume.
- [smolvm sandbox snapshot](https://docs.celesto.ai/smolvm/cli/snapshot.md): Use smolvm sandbox snapshot to save, restore, list, and delete sandbox checkpoints for retries, handoffs, and long-running agent workflows.
- [smolvm sandbox ssh](https://docs.celesto.ai/smolvm/cli/ssh.md): Use smolvm sandbox ssh to open a real SSH session into a sandbox when you need SSH behavior instead of the faster SmolVM shell.
- [smolvm sandbox stop](https://docs.celesto.ai/smolvm/cli/stop.md): Use smolvm sandbox stop to shut down a running sandbox while keeping its record available for later inspection or restart.
- [smolvm ui](https://docs.celesto.ai/smolvm/cli/ui.md): Use smolvm ui to start the local dashboard web server, where you can manage VMs, view logs, and monitor sandbox resources from your browser.
- [smolvm update](https://docs.celesto.ai/smolvm/cli/update.md): Upgrade SmolVM to the latest stable release with one command. Check for updates without installing, or pipe machine-readable JSON output into scripts.
- [smolvm windows](https://docs.celesto.ai/smolvm/cli/windows.md): Build a reusable Windows 11 qcow2 image with smolvm windows build-image — an unattended install that produces a ready-to-boot SmolVM guest.
- [Firecracker, QEMU, and libkrun backends](https://docs.celesto.ai/smolvm/concepts/backends.md): Compare the SmolVM runtime backends: Firecracker for Linux production, QEMU for macOS and compatibility, and experimental libkrun support.
- [Control channel](https://docs.celesto.ai/smolvm/concepts/control-channel.md): How SmolVM runs commands, copies files, and opens shells inside a sandbox over SSH or the faster vsock guest-agent channel.
- [Network configuration](https://docs.celesto.ai/smolvm/concepts/networking.md): How SmolVM networking works — TAP devices, private subnets, NAT, and outbound internet access — so sandboxes stay isolated from each other and the host.
- [SmolVM architecture overview](https://docs.celesto.ai/smolvm/concepts/overview.md): Architecture overview of SmolVM — how each sandbox boots a microVM, sets up isolated networking, runs your code over vsock or SSH, and tears itself down.
- [Published images](https://docs.celesto.ai/smolvm/concepts/published-images.md): How SmolVM boots from pre-built sandbox images so smolvm claude start launches in seconds.
- [Security model](https://docs.celesto.ai/smolvm/concepts/security.md): SmolVM security model: how hardware-virtualized microVMs isolate sandboxes from your host, where the trust boundaries are, and best practices for safe use.
- [Bridged networking](https://docs.celesto.ai/smolvm/features/bridged-networking.md): Attach a Linux sandbox directly to an existing host bridge so it appears as a regular machine on that network, with its own MAC address and guest-managed IP, instead of sitting behind SmolVM's private NAT.
- [Browser sandboxes](https://docs.celesto.ai/smolvm/features/browser-sandboxes.md): Run a real browser inside an isolated SmolVM sandbox for agents, Playwright automation, live viewing, and computer-use workflows.
- [Run callbacks and safety hooks](https://docs.celesto.ai/smolvm/features/callbacks.md): Attach Python callbacks to a SmolVM sandbox to inspect, log, or block commands before they run on the guest — useful for safety filters and audit trails.
- [Coding Agents in Sandbox](https://docs.celesto.ai/smolvm/features/coding-agents.md): Run Claude Code, Codex, Hermes, and Pi coding agents inside a SmolVM sandbox with full permissions — no accept-changes prompts and no host risk.
- [Mount Folders and Data](https://docs.celesto.ai/smolvm/features/host-mounts.md): Share local folders and project files with a SmolVM sandbox using read-only or writable host mounts so agents can explore code without copying it first.
- [Network Controls](https://docs.celesto.ai/smolvm/features/network-controls.md): Restrict sandbox internet access to approved domains using allowed_domains, so agents can reach specific APIs while everything else is blocked at the network.
- [Port Forwarding](https://docs.celesto.ai/smolvm/features/port-forwarding.md): Forward ports from a SmolVM sandbox to your host so you can reach web servers, databases, and APIs running inside the sandbox as if they were local.
- [Snapshot and restore SmolVM sandboxes](https://docs.celesto.ai/smolvm/features/snapshots.md): Snapshot and restore running SmolVM sandboxes to checkpoint agents, preserve configured environments, retry failed steps, and skip cold-start setup.
- [Framework Integration](https://docs.celesto.ai/smolvm/guides/ai-agent-integration.md): Wire SmolVM into PydanticAI, the OpenAI Agents SDK, LangChain, and other agent frameworks so AI agents can run untrusted code inside isolated microVMs.
- [Basic usage](https://docs.celesto.ai/smolvm/guides/basic-usage.md): Day-to-day SmolVM workflows — create sandboxes, run shell commands, check exit codes and output, and handle errors using the Python SDK.
- [Custom Sandbox Images with Docker](https://docs.celesto.ai/smolvm/guides/custom-images.md): Build SmolVM sandboxes from Dockerfiles, SSH-ready Linux images, or existing disk images so your agents start with the tools they need.
- [Environment variables in SmolVM sandboxes](https://docs.celesto.ai/smolvm/guides/environment-variables.md): Inject API keys, secrets, and config into SmolVM sandboxes at boot or at runtime with one Python API that works on Linux and Windows guests.
- [Use the SmolVM HTTP API and TypeScript SDK](https://docs.celesto.ai/smolvm/guides/http-api-typescript-sdk.md): Start the local SmolVM HTTP API, create and manage sandboxes over REST, and use the generated TypeScript Smolvm client from a project that includes the SDK package.
- [VM lifecycle management](https://docs.celesto.ai/smolvm/guides/vm-lifecycle.md): Manage the full SmolVM sandbox lifecycle — created, running, paused, and stopped states — and learn when to use the context manager versus manual control.
- [Windows sandboxes in SmolVM](https://docs.celesto.ai/smolvm/guides/windows-guests.md): Boot a Windows 11 microVM with SmolVM: run PowerShell over SSH, upload files, and inject environment variables from Python on a Linux host.
- [Install SmolVM on Linux or macOS](https://docs.celesto.ai/smolvm/installation.md): Install SmolVM on Linux or macOS using the install script or pip — sets up the Firecracker or QEMU backend automatically based on your platform.
- [SmolVM: secure microVM sandboxes for AI agents](https://docs.celesto.ai/smolvm/introduction.md): Introduction to SmolVM — an open-source microVM sandbox with sub-second boot, hardware isolation, and persistent state for running AI agents safely.
- [Setup Gmail and Calendar integration](https://docs.celesto.ai/superauth/google-oauth-setup.md): Connect Gmail and Google Calendar to your AI agent using the agentor setup-google command — handles OAuth credentials and browser-based authentication.