StackBridge
Sub-1ms cross-stack AST contract layer and compiler verification engine mapping React/Next.js routes to FastAPI and SQLAlchemy models for AI coding agents.
README
<div align="center">
🌉 StackBridge-MCP
Sub-1ms Cross-Stack AST Contract & Verification Layer for AI Coding Agents
<p align="center"> <a href="#-quick-start">Quick Start</a> • <a href="#-client-configuration">Client Config</a> • <a href="#-real-world-benchmarks">Benchmarks</a> • <a href="#-architecture--mcp-tools">MCP Tools</a> • <a href="#-cli-reference">CLI Reference</a> • <a href="docs/architecture.md">Docs</a> </p>
</div>
💡 Why StackBridge?
When AI coding agents (Cursor, Claude Code, Windsurf, Antigravity) edit backend models or API routes in full-stack codebases, backend unit tests frequently pass while the frontend silently breaks in production:
- An agent modifies an API parameter or Pydantic/SQLAlchemy field in
backend/routes.py. - Backend tests pass in isolation. Nothing warns the agent.
- The React/Next.js client calling that endpoint across the boundary fails with runtime errors.
StackBridge-MCP is an always-warm Model Context Protocol (MCP) server that parses full-stack AST relationships, discovers cross-stack blast radii in 0.75 ms, and verifies changes using baseline-diffed compiler checks with zero false positives.
React / Next.js Client FastAPI Routes SQLAlchemy ORM Models
(TypeScript AST) ───► (Python AST) ───► (Schema AST)
UserProfile.tsx get_user_billing() BillingAccount
⚡ Key Highlights
- 🌲 Tree-sitter AST Graph: Parses Next.js (
fetch, Axios, React Query) ↔ FastAPI routes ↔ SQLAlchemy ORM models without heavy LSP sidecars or runtime imports. - ⚡ Sub-1ms Traversal: Persistent SQLite WAL database with recursive Common Table Expressions (0.75 ms traversal query latency).
- 📉 99.74% Prompt Token Reduction: Replaces massive multi-file code dumps with compact, mathematically precise AST contract slices.
- 🛡️ Root-Cause Diagnostic Ranking: Graph-distance BFS ranks errors (
🔴 PRIMARY ROOT CAUSEvs⚠️ CASCADING BREAKAGE) and outputs immediate Git diff patches. - 🧪 Test Impact Selection: Isolates test suites impacted by a schema change and highlights untested blast-radius paths (0% coverage).
- 🌐 Interactive Canvas: Built-in localhost tripartite visualizer (
stackbridge ui) onhttp://127.0.0.1:3456. - 🔄 Continuous Intelligence: Background file watcher daemon (
stackbridge watch) and livingAGENTS.mdcontext generator.
📊 Real-World Benchmarks
Empirical performance measured on fastapi-realworld-example-app (44 files, 23 AST dependency nodes, 10 cross-boundary edges):
| Benchmark Metric | Raw Codebase Dump | StackBridge Compact Slice | Improvement / Latency |
|---|---|---|---|
| Context Window Size | 19,705 tokens |
51 tokens |
📉 99.74% Token Reduction |
| Blast Radius Traversal | Full-repo search: ~150 ms |
SQLite Recursive CTE: 0.75 ms |
⚡ 200x Faster Traversal |
| Compiler Verification | Global linter: ~3,500 ms |
Baseline-Diffed Engine: 312 ms |
🛡️ Zero False Positives |
| Automated Test Suite | — | 56 / 56 tests passing | ✅ 100% Passing |
See full benchmark methodology in docs/benchmarks.md and REAL_WORLD_BENCHMARK.md.
🚀 Quick Start
Option 1: Zero-Install Execution (Recommended via uvx)
uvx stackbridge serve
Option 2: Pip Installation
pip install stackbridge
stackbridge serve
⚙️ Client Configuration
Connect StackBridge to your AI pair programmer over standard JSON-RPC 2.0 stdio:
1. Cursor (.cursor/mcp.json)
{
"mcpServers": {
"stackbridge": {
"command": "uvx",
"args": ["stackbridge", "serve"]
}
}
}
2. Claude Desktop (claude_desktop_config.json)
{
"mcpServers": {
"stackbridge": {
"command": "python",
"args": ["-m", "stackbridge.main", "serve", "--transport", "stdio"]
}
}
}
🤖 MCP Tools Reference
StackBridge exposes high-ergonomics tools to coding agents:
| Tool Name | Arguments | Description |
|---|---|---|
trace_fullstack_path |
symbol_or_path: str |
Traces the full-stack dependency chain: Frontend component ➔ API route ➔ Database model. |
get_route_contract |
route_path: str |
Extracts HTTP methods, status codes, response models, and linked frontend fetch callers with confidence scores. |
verify_schema_change |
modified_files: dict |
Runs in-memory compiler checks across impacted files, ranking root causes and proposing diff patches. |
get_stack_health |
repo_path: str |
Returns real-time full-stack boundary stats, node counts, edge counts, and breakage drift status. |
💻 CLI Reference
# Index a repository and export the dependency graph
stackbridge index --repo-path . --force
# Trace blast radius for a model or route
stackbridge trace --target BillingAccount
# Run pre-commit boundary verification guard
stackbridge guard --fail-on-error
# Launch interactive tripartite web visualizer
stackbridge ui --port 3456
# Start continuous background watcher daemon
stackbridge watch
# Generate living AGENTS.md boundary architecture guide
stackbridge init-agents
# Execute performance and token reduction benchmarks
stackbridge benchmark --runs 3 --output BENCHMARK.md
📁 Repository Structure
StackBridge-MCP/
├── .github/
│ ├── workflows/ci.yml # CI pipeline (Python 3.10-3.13 on Ubuntu/Windows/macOS)
│ ├── ISSUE_TEMPLATE/ # Bug report and feature request issue templates
│ └── PULL_REQUEST_TEMPLATE.md # Standard PR checklist
├── docs/
│ ├── architecture.md # Subsystem breakdown and Mermaid diagrams
│ ├── benchmarks.md # Benchmark methodology and raw metrics
│ └── ast_extraction_spec.md # Tree-sitter extractor grammar specifications
├── stackbridge/
│ ├── core/ # Unified StackGraph, SQLite CTE store, watcher, route matcher
│ ├── parsers/ # Tree-sitter parsers (TS fetch, Python routes, SQLAlchemy)
│ ├── verifier/ # Baseline-diffed verifier, root-cause ranker, test impact selector
│ ├── mcp_server/ # FastMCP stdio server and JSON-RPC tools
│ ├── benchmarks/ # Benchmark runner and markdown report generator
│ └── ui/ # Localhost tripartite interactive canvas
├── tests/ # 56 automated test suites (parsers, verifiers, MCP E2E, CTE)
├── AGENTS.md # Living agent architecture guide
├── CHANGELOG.md # Version release notes
├── CONTRIBUTING.md # Contribution and development guidelines
├── LICENSE # MIT License
└── pyproject.toml # Package metadata and tool configurations
📄 License
This project is licensed under the MIT License.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。