MCP for AI assistants

Aloof exposes a Model Context Protocol server, so assistants can pick plans, deploy, read logs, scale, attach volumes and add domains for you or your users.

Remote (OAuth 2.1): add https://mcp.aloof.si/mcp as a connector in Claude, ChatGPT or Cursor. The assistant registers itself (dynamic client registration), sends you to Aloof to approve, and gets a scoped token (PKCE).

Local (stdio):

{"mcpServers": {"aloof": {"command": "python3", "args": ["mcp_server/server.py"],
  "env": {"ALOOF_API_URL": "https://api.aloof.si", "ALOOF_API_KEY": "pk_..."}}}}

Tools: list_plans, signup, deploy_app, get_status, list_apps, get_logs, scale_app, set_env, destroy_app, get_usage, create_volume, list_volumes, attach_volume, detach_volume, set_autoscale, add_domain, list_releases, rollback_app, create_postgres, list_postgres, attach_postgres, backup_postgres, snapshot_volume, restore_volume, exec_command, add_log_drain, list_offsite_backups, restore_offsite_backup, registry_login, add_git_credential, deploy_with_config, app_status, create_deploy_token.

Discovery: /llms.txt, /.well-known/mcp.json.