OpenSumi
Author: opensumi
Description: A TypeScript-based framework for building AI-native IDE products (web, Electron, cloud). Includes MCP client support (Model Context Protocol) and indicates support for MCP tools via MCP servers. This repository is not an MCP server itself; it is an IDE framework/client that can connect to MCP servers and use their tools.
Stars: 3.6k
Forks: 449
License: MIT License
Category: Open Source
Overview
Installation
yarn installyarn run inityarn run download-extensionyarn run startMY_WORKSPACE={local_path} yarn run startFAQs
Which MCP features besides tools are planned for future OpenSumi releases?
The content does not specify which MCP features beyond tools are planned for future OpenSumi releases. No public roadmap or feature timeline is documented for resources, prompts, sampling, or roots support. Teams evaluating OpenSumi should check the GitHub repository's issues, pull requests, or discussions for community-driven feature requests regarding expanded MCP capabilities.
How does OpenSumi's VS Code extension compatibility (v1.68.0) affect which extensions can be used in a custom IDE build?
OpenSumi's v1.68.0 API compatibility means extensions requiring newer VS Code APIs will fail at runtime or degrade gracefully depending on fallback logic. Extensions published after June 2022 that depend on later API versions won't function correctly. Teams must audit extension manifests for engine compatibility and test critical extensions against OpenSumi's frozen API surface before deployment.
What are the main challenges of adopting OpenSumi for English-speaking teams given its Chinese-language ecosystem?
English-speaking teams face three primary challenges: limited community resources like Stack Overflow answers or tutorial content in English, slower issue resolution due to language barriers in GitHub discussions, and difficulty recruiting developers with prior OpenSumi experience outside China. Additionally, translating internal documentation and training materials becomes an ongoing operational cost as the framework evolves across breaking changes.
Can OpenSumi connect to multiple MCP servers simultaneously, and how does tool routing work across them?
Yes, OpenSumi can connect to multiple MCP servers simultaneously, as confirmed by the use case describing aggregation of tools from GitHub and Context7 servers. Tool routing mechanics are not documented in public materials; teams must inspect the ai-native package source to understand namespace collision handling, tool-list merging, and invocation priority when identical tool names exist across different servers.
How do I configure MCP servers for use with different AI assistants?
MCP server configuration varies by AI assistant. Most clients use JSON files to define server paths, environment variables, and transport protocols like stdio or SSE. Claude Desktop stores config in JSON preferences, while IDEs like Cursor, Windsurf, and VS Code use workspace settings. OpenSumi uses a preferences-based system with undocumented schema requiring source code review.
What are the differences between using Docker and native installations for MCP servers?
Docker installations provide container-based isolation, sandboxing MCP servers from host filesystem and shell access, reducing risks from compromised packages. Native installations offer direct system integration with lower overhead but expose full shell access, creating vulnerabilities like SSH key exposure. Docker requires port mapping and orchestration, suiting production environments, while native installs simplify Linux deployment through package formats without container runtime costs.