skybridge

Name: skybridge
Rating: 82 (1 reviews)
Author: alpic-ai

MCP Server

alpic-ai/skybridge

Skybridge is a full-stack TypeScript framework designed to help developers and AI agents build interactive Model Context Protocol (MCP) apps and ChatGPT applications. It abstracts runtime differences so that a single backend server and React frontend can render rich UI views across multiple AI clients.

2k 132MITUpdated 2026-08-11

Great

View on GitHub Open source file

Overview

The Skybridge framework addresses the low-level complexities of raw MCP and ChatGPT Apps SDKs by providing a robust type-safe environment complete with React hooks, local emulation tooling, hot module reloading, and an integrated development server. By bridging distinct host runtimes under a unified API, it enables developers to create dynamic, stateful UI components that render directly inside AI conversations.

Capabilities

▸Type-safe end-to-end tRPC-style inference from MCP tool definitions to React views
▸Cross-runtime compatibility supporting both ChatGPT Apps SDK and open MCP Apps standards
▸React Query-style hooks and localized state management for AI conversational interfaces
▸Development server featuring a local emulator, hot module replacement, and permanent tunneling

Best for

Building interactive React-based UI components inside AI conversation interfaces like Claude and ChatGPT, Migrating existing standalone MCP servers to support rich graphical frontend views, Developing type-safe fullstack AI applications with shared state synchronization between model and UI

Works with

Claude CodeCursorGitHub CopilotCodex

Evoa Score breakdown= Σ (score × weight)

Task usefulness20%77 → +15.4

実タスクにどれだけ役立つか（機能の豊富さ・用途の明確さ）。 — AIによるcapabilities/use-cases解析

Code quality15%90 → +13.5

実装・指示の品質。 — AIによるSKILL.md/README解析

Maintenance15%100 → +15.0

リポジトリがどれだけ活発に保守されているか。 — GitHub 最終push日時の新しさ

Documentation12%48 → +5.8

ドキュメントの充実度・分かりやすさ。 — README/独自要約の情報量

Security15%80 → +12.0

危険・不審な挙動が無いか。 — AIによるセキュリティレビュー

Originality10%85 → +8.5

ありふれたラッパーではない独自性。 — AIによる独自性判定

Popularity8%84 → +6.7

コミュニティの採用度。 — GitHub Stars/Forks（対数スケール）

Compatibility5%100 → +5.0

対応AIエージェントの広さ。 — AIによる対応エージェント判定

Weighted total81.9 / 100

ライセンス不明/制限あり（Red）のSkillは総合スコアに0.85倍の補正を適用します。ランキングはこのScoreのみで決まり、広告で変わりません。算出方法の詳細 →

Security considerations

The framework interacts directly with external AI host runtimes via iframes and postMessage protocols, requiring careful management of input validation and state synchronization to prevent injection vectors.