# Agno: AgentOS

## AgentOS

### Get Started

- [What is AgentOS?](https://docs.agno.com/agent-os/introduction.md): Run agents, teams, and workflows using FastAPI.
- [Run Your First AgentOS](https://docs.agno.com/agent-os/run-your-os.md): Serve an agent through FastAPI with persistent sessions and a local REST API.
- [Connect Your AgentOS](https://docs.agno.com/agent-os/connect-your-os.md): Connect an AgentOS runtime to the Control Plane to manage and monitor it from one web interface.
- [AgentOS Control Plane](https://docs.agno.com/agent-os/control-plane.md): Manage and monitor AgentOS runtimes from one web interface.
- [Security & Auth](https://docs.agno.com/agent-os/security/overview.md): Authentication modes, service account tokens, and the single auth layer covering every AgentOS surface.

### AgentOS

- [AgentOS Runtime](https://docs.agno.com/agent-os/overview.md): Configure the FastAPI runtime that serves your agents, teams, and workflows.
- [Using the API](https://docs.agno.com/agent-os/using-the-api.md): Run agents, manage state, and operate AgentOS through its REST API.
- [AgentOS Configuration](https://docs.agno.com/agent-os/config.md): Configure quick prompts, display names, and per-database settings with a YAML file or the AgentOSConfig class
- [Background Hooks](https://docs.agno.com/agent-os/background-tasks/overview.md): Run agent hooks as non-blocking background tasks in AgentOS
- [Scheduler](https://docs.agno.com/agent-os/scheduler/overview.md): Deploy and manage scheduled execution for agents and workflows via AgentOS cron jobs.
- [Approvals](https://docs.agno.com/agent-os/approvals/overview.md): Manage approval workflows for agents and teams via the AgentOS Control Plane.
- [Custom Lifespan](https://docs.agno.com/agent-os/lifespan.md): Customize the lifespan of your AgentOS app to handle startup and shutdown logic.

#### MCP

- [AgentOS as MCP Server](https://docs.agno.com/agent-os/mcp/mcp.md): Serve AgentOS as an MCP server at /mcp, with eight built-in tools, custom tools, and per-call authorization via MCPConfig.
- [MCPTools within AgentOS](https://docs.agno.com/agent-os/mcp/tools.md): Give the Agents, Teams, and Workflows in your AgentOS access to external MCP servers with MCPTools.

#### Interfaces

- [Interfaces](https://docs.agno.com/agent-os/interfaces/overview.md): Expose agents, teams, and workflows through AI frontends, messaging platforms, and agent protocols.
- [A2A](https://docs.agno.com/agent-os/interfaces/a2a/introduction.md): Expose Agno agents via the A2A protocol
- [AG-UI](https://docs.agno.com/agent-os/interfaces/ag-ui/introduction.md): Expose Agno agents via the AG-UI protocol

##### Slack

- [Slack](https://docs.agno.com/agent-os/interfaces/slack/introduction.md): Run agents, teams, and workflows in Slack with session routing, streaming, file handling, and human-in-the-loop approvals.
- [Setup](https://docs.agno.com/agent-os/interfaces/slack/setup.md): Create a Slack App and connect it to your Agno agent.
- [Features](https://docs.agno.com/agent-os/interfaces/slack/features.md): Memory, files, streaming, search, and cross-platform identity for your Slack agent.
- [Human-in-the-Loop](https://docs.agno.com/agent-os/interfaces/slack/hitl.md): Pause agents for approval before executing sensitive tools.
- [Slack Reference](https://docs.agno.com/agent-os/interfaces/slack/reference.md): Interface parameters, endpoints, event handling, and OAuth scopes for the Slack interface.

##### Telegram

- [Telegram](https://docs.agno.com/agent-os/interfaces/telegram/introduction.md): Expose agents, teams, or workflows as Telegram bots with webhook endpoints.
- [Setup](https://docs.agno.com/agent-os/interfaces/telegram/setup.md): Create a Telegram bot with BotFather and configure webhooks for local and production deployments.
- [Telegram Reference](https://docs.agno.com/agent-os/interfaces/telegram/reference.md): Interface parameters, endpoints, and event handling for the Telegram interface.

##### WhatsApp

- [WhatsApp](https://docs.agno.com/agent-os/interfaces/whatsapp/introduction.md): Deploy agents, teams, or workflows as WhatsApp bots via the WhatsApp Interface using Agno.
- [Setup](https://docs.agno.com/agent-os/interfaces/whatsapp/setup.md): Configure Meta Developer account, WhatsApp Business API, and webhooks for WhatsApp bots.
- [WhatsApp Reference](https://docs.agno.com/agent-os/interfaces/whatsapp/reference.md): Interface parameters, endpoints, and webhook handling for the WhatsApp interface.

##### Discord

- [Discord](https://docs.agno.com/agent-os/interfaces/discord/introduction.md): Deploy agents to Discord for community support and moderation.

#### BYO FastAPI App

- [Bring Your Own FastAPI App](https://docs.agno.com/agent-os/custom-fastapi/overview.md): Integrate your own FastAPI app with AgentOS.
- [Overriding Routes](https://docs.agno.com/agent-os/custom-fastapi/override-routes.md): Resolve route conflicts between your custom FastAPI app and AgentOS with the on_route_conflict parameter.

#### Middleware

- [AgentOS Middleware](https://docs.agno.com/agent-os/middleware/overview.md): Add authentication, logging, monitoring, and security features to your AgentOS application using middleware
- [Custom Middleware](https://docs.agno.com/agent-os/middleware/custom.md): Create custom middleware for rate limiting, logging, security, and monitoring in AgentOS
- [Auth Middleware](https://docs.agno.com/agent-os/middleware/jwt.md): Configure AuthMiddleware for JWT validation, claim injection, and RBAC across REST, MCP, and WebSocket connections.

#### Authorization

- [Authorization](https://docs.agno.com/agent-os/security/authorization/overview.md): JWT validation and scope-based permissions for AgentOS endpoints.
- [Quickstart](https://docs.agno.com/agent-os/security/authorization/quickstart.md): Enable authorization, set a verification key, and make your first authenticated request.
- [JSON Web Tokens (JWT)](https://docs.agno.com/agent-os/security/authorization/tokens.md): JWT claim structure, example tokens, and how AgentOS reads them.
- [Self-Hosted (BYO Token)](https://docs.agno.com/agent-os/security/authorization/self-hosted.md): Run AgentOS without the AgentOS control plane by issuing and verifying your own JWTs.
- [Service Accounts](https://docs.agno.com/agent-os/security/authorization/service-accounts.md): Mint, scope, and revoke the opaque agno_pat_ tokens that machine identities use to authenticate with AgentOS.
- [Scopes](https://docs.agno.com/agent-os/security/authorization/scopes.md): Scope format and the full permission reference for every AgentOS endpoint.
- [Roles](https://docs.agno.com/agent-os/security/authorization/roles.md): Default and custom roles in the AgentOS control plane.
- [Per-User Data Isolation](https://docs.agno.com/agent-os/security/authorization/user-isolation.md): Scope user-owned AgentOS data and resources to the caller's user ID.

#### Dynamic Agents

- [Dynamic Agents](https://docs.agno.com/agent-os/factories/overview.md): Build agents, teams, and workflows per request from JWT claims, user input, and other request-time context.
- [AgentFactory](https://docs.agno.com/agent-os/factories/agent-factory.md): Build an Agent per request from verified middleware claims, client input, or any request-time context.
- [TeamFactory](https://docs.agno.com/agent-os/factories/team-factory.md): Build a Team per request whose members, model, and instructions depend on the caller.
- [WorkflowFactory](https://docs.agno.com/agent-os/factories/workflow-factory.md): Build a Workflow per request whose steps, agents, and model depend on the caller.

#### Knowledge

- [Manage Knowledge](https://docs.agno.com/agent-os/knowledge/manage-knowledge.md): Attach Knowledge to your AgentOS instance
- [Remote Content](https://docs.agno.com/agent-os/knowledge/remote-content.md): Upload content from registered cloud sources via the AgentOS Knowledge API.
- [Filter Knowledge](https://docs.agno.com/agent-os/knowledge/filter-knowledge.md): Use filter expressions through the AgentOS API for precise knowledge base filtering.

#### Learnings

- [Manage Learnings](https://docs.agno.com/agent-os/learnings/manage-learnings.md): Read and manage learning records through the AgentOS /learnings endpoints.

#### Studio

- [Overview](https://docs.agno.com/agent-os/studio/introduction.md): A visual editor in AgentOS to build Agents, Teams, and Workflows.
- [Agents](https://docs.agno.com/agent-os/studio/agents.md): Build and configure agents visually in AgentOS Studio.
- [Teams](https://docs.agno.com/agent-os/studio/teams.md): Compose multi-agent teams visually in AgentOS Studio.
- [Registry](https://docs.agno.com/agent-os/studio/registry.md): Register tools, models, databases, workflows, and learning machines for use in AgentOS Studio.
- [StudioTools](https://docs.agno.com/agent-os/studio/tools.md): Give an agent tools to create, update, and run Studio components from chat with human-in-the-loop confirmation.

##### Workflows

- [Overview](https://docs.agno.com/agent-os/studio/workflows.md): Design step-based workflows visually in AgentOS Studio.
- [CEL Expressions](https://docs.agno.com/agent-os/studio/cel-expressions.md): Use CEL expressions as evaluators, end conditions, and selectors in workflow steps.

#### Multi-Framework

- [Multi-Framework Support](https://docs.agno.com/agent-os/multi-framework/overview.md): Serve agents built with the Claude Agent SDK, LangGraph, DSPy and Antigravity from one AgentOS.
- [Antigravity](https://docs.agno.com/agent-os/multi-framework/antigravity.md): Run Google's Managed Agents (Gemini API) as an AgentOS agent using AntigravityAgent.
- [Claude Agent SDK](https://docs.agno.com/agent-os/multi-framework/claude-agent-sdk.md): Run Claude Code as an AgentOS agent using ClaudeAgent.
- [LangGraph](https://docs.agno.com/agent-os/multi-framework/langgraph.md): Wrap a compiled LangGraph graph as an AgentOS agent.
- [DSPy](https://docs.agno.com/agent-os/multi-framework/dspy.md): Serve a DSPy program as an AgentOS agent.

#### Background Execution

- [Background Execution](https://docs.agno.com/agent-os/background-execution/overview.md): Accept runs with background=true, poll or stream them, and make acceptance durable with a database-backed job queue.
- [Durable Queue](https://docs.agno.com/agent-os/background-execution/durable-queue.md): QueueConfig(durable=True): accepted background runs become committed rows that survive crashes and deploys.
- [Multi-Replica Deployments](https://docs.agno.com/agent-os/background-execution/multi-replica.md): QueueConfig(redis=...) sets the event stream and cancellation manager on every replica so runs can be resumed and cancelled from any of them.
- [Human-in-the-Loop Continuations](https://docs.agno.com/agent-os/background-execution/hitl-continuations.md): Continue a paused durable run through the queue with background=true. Same job, same run_id, one more attempt.
- [Operations and Monitoring](https://docs.agno.com/agent-os/background-execution/operations.md): The /queue endpoints: dead-letter listing, requeue, queue stats, retention, admin gating, and deployment affinity.

#### Clients

- [Clients](https://docs.agno.com/agent-os/client/overview.md): Python clients for connecting to agent servers
- [AgentOS Client](https://docs.agno.com/agent-os/client/agentos-client.md): Connect to Agno AgentOS instances via REST API
- [A2A Client](https://docs.agno.com/agent-os/client/a2a-client.md): Connect to any A2A-compatible agent server

#### Remote Execution

- [Remote Execution](https://docs.agno.com/agent-os/remote-execution/overview.md): Execute agents, teams, and workflows hosted on remote AgentOS instances
- [Remote Agent](https://docs.agno.com/agent-os/remote-execution/remote-agent.md): Execute agents hosted on remote AgentOS instances
- [Remote Team](https://docs.agno.com/agent-os/remote-execution/remote-team.md): Execute teams hosted on remote AgentOS instances
- [Remote Workflow](https://docs.agno.com/agent-os/remote-execution/remote-workflow.md): Execute workflows hosted on remote AgentOS instances
- [AgentOS Gateway](https://docs.agno.com/agent-os/remote-execution/gateway.md): Create a unified API gateway for multiple AgentOS instances

#### Tracing

- [Tracing](https://docs.agno.com/agent-os/tracing/overview.md): Store AgentOS traces to inspect run behavior, latency, errors, model calls, and tool calls.
- [Filter Options](https://docs.agno.com/agent-os/tracing/filter-options.md): Filter and search traces using structured queries, time ranges, and view toggles.

##### Usage

- [Basic Agent Tracing](https://docs.agno.com/agent-os/tracing/usage/basic-agent-tracing.md): Trace agents with Agno in AgentOS.
- [Basic Team Tracing](https://docs.agno.com/agent-os/tracing/usage/basic-team-tracing.md): Trace teams with Agno in AgentOS.
- [Basic Workflow Tracing](https://docs.agno.com/agent-os/tracing/usage/basic-workflow-tracing.md): Trace workflows with Agno in AgentOS.
- [Agent with Knowledge Tracing](https://docs.agno.com/agent-os/tracing/usage/agent-with-knowledge-tracing.md): Trace agents with knowledge bases in AgentOS.
- [Agent with Reasoning Tools Tracing](https://docs.agno.com/agent-os/tracing/usage/agent-with-reasoning-tools-tracing.md): Trace agents with reasoning tools in AgentOS.
- [Multi-DB Tracing with setup_tracing()](https://docs.agno.com/agent-os/tracing/usage/tracing-with-multi-db-scenario.md): Trace agents with multiple databases using setup_tracing() in AgentOS.
- [Multi-DB Tracing with tracing=True](https://docs.agno.com/agent-os/tracing/usage/tracing-with-multi-db-scenario-and-tracing-flag.md): Trace agents with multiple databases using tracing=True in AgentOS.

### Example Usage

- [AgentOS Demo](https://docs.agno.com/agent-os/usage/demo.md): AgentOS with a knowledge-backed agent, a web research team, and Postgres storage.
- [AgentOS Configuration](https://docs.agno.com/agent-os/usage/extra-configuration.md): Pass a YAML configuration file to AgentOS for quick prompts and database display names.
- [Human-in-the-Loop Example](https://docs.agno.com/agent-os/usage/hitl.md): AgentOS with tools requiring user confirmation
- [Database Migrations](https://docs.agno.com/agent-os/usage/database-migrations.md): Migrate your AgentOS database schema.

#### MCP

- [Enable AgentOS MCP](https://docs.agno.com/agent-os/usage/mcp/enable-mcp-example.md): AgentOS with the MCP server enabled, plus an agent client that authenticates and operates it.
- [AgentOS with MCPTools](https://docs.agno.com/agent-os/usage/mcp/mcp-tools-example.md): AgentOS whose agent connects to an external MCP server through MCPTools.

#### Middleware

- [Custom Middleware](https://docs.agno.com/agent-os/usage/middleware/custom-middleware.md): AgentOS with custom middleware for rate limiting and request logging
- [JWT Middleware with Authorization Headers](https://docs.agno.com/agent-os/usage/middleware/jwt-middleware.md): Complete AgentOS setup with JWT middleware for authentication and parameter injection using Authorization headers
- [JWT Middleware with Cookies](https://docs.agno.com/agent-os/usage/middleware/jwt-cookies.md): AgentOS with JWT middleware using HTTP-only cookies for secure web authentication
- [Custom FastAPI App with JWT Middleware](https://docs.agno.com/agent-os/usage/middleware/custom-fastapi-jwt.md): Custom FastAPI application with JWT middleware for authentication and AgentOS integration

#### RBAC

- [Basic RBAC (Symmetric)](https://docs.agno.com/agent-os/usage/rbac/basic-symmetric.md): AgentOS with RBAC enabled using JWT HS256 and a shared secret.
- [Basic RBAC (Asymmetric)](https://docs.agno.com/agent-os/usage/rbac/basic-asymmetric.md): Enable RBAC with RS256 asymmetric JWT keys.
- [WorkOS BYOT (Bring Your Own Token)](https://docs.agno.com/agent-os/usage/rbac/workos-byot.md): Verify WorkOS-issued JWTs and read scopes from the permissions claim.
- [Advanced Scopes](https://docs.agno.com/agent-os/usage/rbac/advanced-scopes.md): Use global, per-resource, and wildcard RBAC scope patterns.
- [Per-Agent Permissions](https://docs.agno.com/agent-os/usage/rbac/per-agent-permissions.md): Restrict which users can run which agents with per-agent scopes.
- [Custom Scope Mappings](https://docs.agno.com/agent-os/usage/rbac/custom-scope-mappings.md): Map custom scopes to AgentOS routes instead of the defaults.

#### Background Tasks

- [Background Hooks (Global)](https://docs.agno.com/agent-os/usage/background-hooks-global.md): Run non-guardrail agent hooks as background tasks using AgentOS
- [Background Hooks (Per-Hook)](https://docs.agno.com/agent-os/usage/background-hooks-decorator.md): Run specific hooks as background tasks using the @hook decorator
- [Background Output Evaluation](https://docs.agno.com/agent-os/usage/background-output-evaluation.md): Use Agent as Judge evaluation to assess responses as a background task

#### Client

- [Basic Client Usage](https://docs.agno.com/agent-os/usage/client/basic-client.md): Connect to an AgentOS instance and inspect its agents, teams, and workflows
- [Running Agents](https://docs.agno.com/agent-os/usage/client/run-agents.md): Execute agent runs with streaming, non-streaming, and resumable background responses.
- [Running Teams](https://docs.agno.com/agent-os/usage/client/run-teams.md): Execute team runs with streaming and non-streaming responses
- [Running Workflows](https://docs.agno.com/agent-os/usage/client/run-workflows.md): Execute workflow runs with streaming and non-streaming responses
- [Session Management](https://docs.agno.com/agent-os/usage/client/session-management.md): Create, list, and manage sessions for agents, teams, and workflows
- [Memory Operations](https://docs.agno.com/agent-os/usage/client/memory-operations.md): Create, update, list, and delete user memories
- [Knowledge Search](https://docs.agno.com/agent-os/usage/client/knowledge-search.md): Search the knowledge base and list stored content

#### Remote Execution

- [Remote Agent](https://docs.agno.com/agent-os/usage/remote-execution/remote-agent.md): Execute agents hosted on a remote AgentOS instance
- [Remote Team](https://docs.agno.com/agent-os/usage/remote-execution/remote-team.md): Execute teams hosted on a remote AgentOS instance
- [AgentOS Gateway](https://docs.agno.com/agent-os/usage/remote-execution/gateway.md): Create a unified API gateway for multiple AgentOS instances

#### Interfaces

##### WhatsApp

- [Basic WhatsApp Agent](https://docs.agno.com/agent-os/usage/interfaces/whatsapp/basic.md): Create a basic AI agent that integrates with the WhatsApp Business API
- [WhatsApp Image Generation Agent (Tool-based)](https://docs.agno.com/agent-os/usage/interfaces/whatsapp/image-generation-tools.md): WhatsApp agent that generates images using OpenAI's image generation tools
- [WhatsApp Reasoning Finance Agent](https://docs.agno.com/agent-os/usage/interfaces/whatsapp/reasoning-agent.md): WhatsApp agent with advanced reasoning and financial analysis capabilities
- [WhatsApp Image Generation Agent (Model-based)](https://docs.agno.com/agent-os/usage/interfaces/whatsapp/image-generation-model.md): WhatsApp agent that generates images using Gemini's built-in capabilities
- [WhatsApp Agent with Media Support](https://docs.agno.com/agent-os/usage/interfaces/whatsapp/agent-with-media.md): WhatsApp agent that analyzes images, videos, and audio using multimodal AI
- [WhatsApp Agent with User Memory](https://docs.agno.com/agent-os/usage/interfaces/whatsapp/agent-with-user-memory.md): Personalized WhatsApp agent that remembers user information and preferences

##### A2A

- [Basic](https://docs.agno.com/agent-os/usage/interfaces/a2a/basic.md): Create a basic AI agent with A2A interface
- [Agent with Tools](https://docs.agno.com/agent-os/usage/interfaces/a2a/agent-with-tools.md): Investment analyst agent with financial tools served over the A2A protocol
- [Research Team](https://docs.agno.com/agent-os/usage/interfaces/a2a/team.md): Multi-agent research team with specialized roles served over the A2A protocol

##### AG-UI

- [Basic](https://docs.agno.com/agent-os/usage/interfaces/ag-ui/basic.md): Expose an OpenAI agent through the AG-UI protocol.
- [Agent with Tools](https://docs.agno.com/agent-os/usage/interfaces/ag-ui/agent-with-tools.md): Investment analyst agent with financial tools served over the AG-UI protocol
- [Research Team](https://docs.agno.com/agent-os/usage/interfaces/ag-ui/team.md): Multi-agent research team with specialized roles and web interface

##### Slack

- [Basic Slack Agent](https://docs.agno.com/agent-os/usage/interfaces/slack/basic.md): Minimal Slack bot with SQLite session persistence and @mention replies
- [Streaming Deep Research](https://docs.agno.com/agent-os/usage/interfaces/slack/streaming.md): Multi-tool research agent with streaming task cards and suggested prompts
- [Support Team](https://docs.agno.com/agent-os/usage/interfaces/slack/support-team.md): Multi-agent team with SlackTools that routes support questions to the right specialist.
- [Channel Summarizer](https://docs.agno.com/agent-os/usage/interfaces/slack/channel-summarizer.md): SlackTools-powered agent that reads channel history and produces structured summaries
- [Slack Workflow](https://docs.agno.com/agent-os/usage/interfaces/slack/workflow.md): Research and write two-step workflow with WebSearch and SQLite sessions
- [Multiple Instances](https://docs.agno.com/agent-os/usage/interfaces/slack/multiple-instances.md): Multiple Slack bots on a single AgentOS server with prefix-based routing
- [Slack Agent with User Memory](https://docs.agno.com/agent-os/usage/interfaces/slack/agent-with-user-memory.md): MemoryManager captures user preferences and personalizes responses across conversations
- [Reasoning Finance Agent](https://docs.agno.com/agent-os/usage/interfaces/slack/reasoning-agent.md): ReasoningTools and WebSearch for step-by-step financial analysis on Slack
- [File Analyst](https://docs.agno.com/agent-os/usage/interfaces/slack/file-analyst.md): Slack bot that downloads shared files, analyzes their content, and uploads results back to the channel
- [Multi-Bot](https://docs.agno.com/agent-os/usage/interfaces/slack/multi-bot.md): Two independent Slack bots on the same workspace with separate prefixes, tokens, and session isolation
- [Multimodal Team](https://docs.agno.com/agent-os/usage/interfaces/slack/multimodal-team.md): Legacy Slack multimodal team with vision analysis and deprecated DALL-E image generation
- [Multimodal Workflow](https://docs.agno.com/agent-os/usage/interfaces/slack/multimodal-workflow.md): Legacy Slack workflow with visual analysis, web research, and deprecated DALL-E image generation
- [Research Assistant](https://docs.agno.com/agent-os/usage/interfaces/slack/research-assistant.md): Agent that searches Slack message history and the web to answer research questions

##### Telegram

- [Basic Telegram Agent](https://docs.agno.com/agent-os/usage/interfaces/telegram/basic.md): Gemini agent with session persistence on Telegram
- [Streaming Telegram Agent](https://docs.agno.com/agent-os/usage/interfaces/telegram/streaming.md): OpenAI agent with token-by-token streaming via live message edits
- [Multi-Agent Telegram Team](https://docs.agno.com/agent-os/usage/interfaces/telegram/team.md): Researcher + Writer team coordinating on Telegram
- [Telegram Agent with Media](https://docs.agno.com/agent-os/usage/interfaces/telegram/agent-with-media.md): Legacy Telegram media bot with deprecated DALL-E image generation and ElevenLabs TTS
- [Telegram Agent with User Memory](https://docs.agno.com/agent-os/usage/interfaces/telegram/agent-with-user-memory.md): MemoryManager for cross-session user recall on Telegram
- [Telegram Reasoning Agent](https://docs.agno.com/agent-os/usage/interfaces/telegram/reasoning-agent.md): ReasoningTools + DuckDuckGo search on Telegram
- [Telegram Workflow](https://docs.agno.com/agent-os/usage/interfaces/telegram/workflow.md): Draft + Edit two-step workflow on Telegram
- [Streaming Workflow](https://docs.agno.com/agent-os/usage/interfaces/telegram/streaming-workflow.md): Research + Write workflow with live step progress on Telegram.
- [Multiple Instances](https://docs.agno.com/agent-os/usage/interfaces/telegram/multiple-instances.md): Multiple Telegram bots on a single AgentOS server