feat: Daytona remote workspace implemented

[brunogrbavac]

TLDR

Added Daytona implementation support as a new remote workspace type.
Daytona provides a secure and elastic infrastructure for running AI-generated code.

What changed?

• Added Daytona workspace implementation: New workspace type using Daytona's sandbox environments
• Created workspace infrastructure: New DaytonaWorkspace class with setup/teardown lifecycle
• Sandbox configuration options: api_key, api_url, target, language, snapshot, env_vars, labels, public, timeout, auto_stop_interval, auto_archive_interval, auto_delete_interval, and os_user parameters
• Workspace setup: Sandbox creation from snapshots, preview token extraction, background Composio server startup, and async session management
• Updated documentation: Added Daytona workspace section with examples
• Added CI/CD integration: DAYTONA_API_KEY_STAGING to GitHub Actions and tox
• Added dependencies: daytona>=0.22.1 to setup.py and tox.ini

How to test?

from composio import ComposioToolSet, WorkspaceType

# Simple usage
toolset = ComposioToolSet(workspace_config=WorkspaceType.Daytona())

# With custom configuration
toolset = ComposioToolSet(workspace_config=WorkspaceType.Daytona(
    snapshot="daytona-composio",
    api_key="your_daytona_api_key", 
    api_url="https://app.daytona.io/api",
    target="eu"
))

# Run tests
export DAYTONA_API_KEY="your_api_key"
tox -e test -- -m 'not e2e and not swe'

Why make this change?

This adds Daytona as a fifth workspace option alongside existing Host, Docker, E2B, and FlyIO environments. The snapshot-based approach provides:

• Consistent environments using predefined snapshots
• Faster startup times compared to other remote options
• Better error handling and resource management
• Integration with existing Composio toolset architecture

[entelligence-ai-pr-reviews]

LGTM 👍