---
id: composio
version: "0.19.0"
license: unclear
license_treatment: permissive
maintenance: active
---

# composio — SDK for integrating Composio with your applications.

License: permissive · Maintenance: active · Downloads: 1.9M/mo

## What it is and what it does

Composio is a Python SDK that abstracts away the complexity of connecting AI agents to external applications. Instead of manually building OAuth flows, managing API credentials, or writing tool adapters for each framework, you create a per-user session and retrieve pre-built tool definitions in your agent framework's native format. The SDK handles authentication, session persistence, and runtime tool discovery through meta tools—lightweight abstractions that let agents discover and execute app actions without loading hundreds of tool definitions upfront.

The package integrates with major AI frameworks via separate provider packages. It also exposes a hosted MCP endpoint for Claude, Cursor, and other MCP clients. For event-driven workflows, it supports subscribing to triggers from connected apps and handling webhooks in production.

Use it for:

- Build an AI assistant that reads emails, manages calendars, or sends messages without writing OAuth code.
- Create a multi-tenant SaaS where each user connects their own app accounts within a sandboxed session.
- Set up event-driven automation: subscribe to app events and trigger agent workflows in response.
- Integrate Composio tools into agent pipelines to give agents access to pre-authenticated actions.
- Expose agent capabilities to Claude or Cursor via MCP without managing separate tool registration.
- Prototype multi-step workflows across apps without building custom integrations.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Composio provides a Python SDK that connects AI agents to 1000+ pre-authenticated third-party app toolkits, managing per-user sessions, OAuth flows, event triggers, and sandboxed execution.

Yes. Composio is actively maintained, has low install friction, carries no known vulnerabilities, and solves a real pain point: connecting AI agents to external apps without manual OAuth and credential management. It's worth installing if you're building agents that need to act on third-party services. The permissive Apache license removes legal friction. Primary constraint: requires Python 3.10+ and a Composio API key.

## Install

pip install composio
uv add composio
poetry add composio

## Installing composio

Before you install: Low install friction with a pure-Python wheel. Active maintenance: released 3 days ago with 29688 repository stars and a last commit on 2026-08-14. Requires Python 3.10+. Seven runtime dependencies including pydantic, openai, and composio-client.

License in practice: Licensed under Apache Software License (permissive), allowing commercial use, modification, and distribution with minimal restrictions.

Quickstart:

pip install composio

from composio import Composio

composio = Composio()  # reads COMPOSIO_API_KEY
session = composio.create(user_id="user_123")
tools = session.tools()  # OpenAI-format tool definitions

Requires Python 3.10 or later and a valid COMPOSIO_API_KEY environment variable or explicit api_key parameter.

Verify before relying:

- Whether all 1000+ toolkits are available in every region or account tier.
- Rate limits or quota constraints on tool execution or session creation.
- Latency characteristics for meta-tool discovery and authentication flows.
- Data residency or compliance certifications for the hosted MCP endpoint.

## Package facts

- License: not declared (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 1.9M/month (top 5,000 on PyPI)
- Known vulnerabilities: none known

## Tags

ai agent tool integration, pre-authenticated app toolkits, agent framework integration, oauth session management, ai agent actions, composio sdk, agent tool orchestration, ai-agent-tools, oauth-integration, multi-tenant-saas

[View on SkillFed](https://skillfed.io/packages/composio) · [View on PyPI](https://pypi.org/project/composio/)