by AriaScarlet132
Agent Data Bridge: secure data integration server bridging Spring Boot interfaces with a lightweight Python sandbox—expo
Connects agents to Spring Boot data sources and provides a Python sandbox for data analysis. Fetches data in Markdown or Parquet formats for automated processing.
Agent Data Bridge is a community-built MCP server published by AriaScarlet132 that provides AI assistants with tools and capabilities via the Model Context Protocol. Agent Data Bridge: secure data integration server bridging Spring Boot interfaces with a lightweight Python sandbox—expo It is categorized under developer tools, analytics data.
You can install Agent Data Bridge in your AI client of choice. Use the install panel on this page to get one-click setup for Cursor, Claude Desktop, VS Code, and other MCP-compatible clients. This server runs locally on your machine via the stdio transport.
MIT
Agent Data Bridge is released under the MIT license. This is a permissive open-source license, meaning you can freely use, modify, and distribute the software.
Add new capabilities to Claude beyond text generation
Example
Access external data sources, execute code, interact with tools and services
Transform Claude from chatbot to action-taking agent
Provide Claude with access to relevant context and data
Example
Load project documentation, access knowledge bases, query databases
Get more accurate, context-aware responses
Automate multi-step workflows combining AI and external tools
Example
Research → Summarize → Create document → Send notification
Complete complex tasks end-to-end without manual steps
Share your MCP server with the developer community
Agent Data Bridge is among the better-indexed MCP projects we tried; the explainx.ai summary tracks the official description.
We wired Agent Data Bridge into a staging workspace; the listing’s GitHub and npm pointers saved time versus hunting across READMEs.
Agent Data Bridge is among the better-indexed MCP projects we tried; the explainx.ai summary tracks the official description.
Agent Data Bridge reduced integration guesswork — categories and install configs on the listing matched the upstream repo.
We wired Agent Data Bridge into a staging workspace; the listing’s GitHub and npm pointers saved time versus hunting across READMEs.
We evaluated Agent Data Bridge against two servers with overlapping tools; this profile had the clearer scope statement.
Agent Data Bridge reduced integration guesswork — categories and install configs on the listing matched the upstream repo.
Strong directory entry: Agent Data Bridge surfaces stars and publisher context so we could sanity-check maintenance before adopting.
We evaluated Agent Data Bridge against two servers with overlapping tools; this profile had the clearer scope statement.
Useful MCP listing: Agent Data Bridge is the kind of server we cite when onboarding engineers to host + tool permissions.
showing 1-10 of 66
一个面向 Agent 的“数据桥接 + 轻量沙盒执行”服务,提供两种对外形态:
目前实际实现包含:
uv syncuv run -- uvicorn app.main:app --reload --app-dir src默认监听 http://127.0.0.1:8000。
如果希望“一条命令同时启动 REST + MCP”,可使用:
uv run -- python -m app.run_all1.(可选)启动 MCP Server(SSE)
本项目的 MCP Server 默认监听 0.0.0.0:9000(可通过 MCP_HOST/MCP_PORT 修改),并使用 SSE 传输:
GET /ssePOST /messages/启动命令:
$env:PYTHONPATH = "src"
uv run -- python -m app.mcp_server
src 目录运行(一次性):pushd src
uv run -- python -m app.mcp_server
popd
PYTHONPATH=src uv run -- python -m app.mcp_server
复制 .env.example 为 .env 后按需修改。
当前代码实际会用到的配置(与 .env.example 保持一致):
REST_HOST / REST_PORT:REST(FastAPI) 监听地址(用于 Docker 启动与一键启动脚本)。MCP_HOST / MCP_PORT:MCP(SSE) 监听地址。APP_ID / APP_SECRET:Spring Boot OAuth2 client credentials(默认 agent/agent)。说明:
SPRING_BOOT_BASE_URL / SPRING_BOOT_API_PATH 目前在代码中未被使用;/api/fetch 与 MCP 的 fetch_data 都会直接使用传入的 host 参数作为目标地址(见 src/app/services/springboot_client.py)。GET /health返回:{"status":"ok"}
POST /api/fetch请求体:
{
"host": "http://192.168.10.21:3000",
"userid": "Admin",
"sql": "select ...",
"dataset": "demo"
}
行为(与实现一致,见 src/app/main.py):
data.markdown(Markdown 表格)。Rows <= 15:直接返回完整 Markdown 表格。Rows > 15:保存为 parquet 到 SANDBOX_DIR,并返回字段预览 + 前 5 行。响应:
{ "message": "..." }
示例(curl):
curl -X POST http://127.0.0.1:8000/api/fetch \
-H "Content-Type: application/json" \
-d '{"host":"http://192.168.10.21:3000","userid":"Admin","sql":"select 1","dataset":"demo"}'
POST /api/sandbox/run方式 A:JSON
{ "filename": "anything.py", "code": "print(123)" }
方式 B:multipart/form-data
file(.py 文件)code / filename返回(与实现一致,见 src/app/services/sandbox.py):
{
"filename": "script_xxx.py",
"exit_code": 0,
"stdout": "...",
"stderr": "..."
}
MCP Server 目前提供以下 tools(见 src/app/mcp_server.py):
fetch_data(host, userid, sql, dataset) -> strsandbox_run(code, filename=None) -> dictsandbox_list_files() -> str常见用法:先 fetch_data 生成 parquet 文件名,再用 sandbox_run 执行 Python 读取:
import pandas as pd
df = pd.read_parquet("<file_name>")
print(df.head())
sandbox_storage/:默认沙盒数据目录(可通过 SANDBOX_DIR 覆盖)。sandbox_storage/_scripts/:沙盒执行时写入的临时脚本目录(自动创建)。SANDBOX_DIR,并加了超时与输出截断;请勿在不可信输入场景直接暴露到公网。client_id/client_secret 已改为从 .env 读取(APP_ID/APP_SECRET,默认 agent/agent)。同时启动 REST + MCP(推荐使用 compose,并把沙盒目录挂载到宿主机):
docker compose up --build
端口:
http://127.0.0.1:${REST_PORT:-8000}http://127.0.0.1:${MCP_PORT:-9000}/sse数据卷:
./sandbox_storage -> /app/sandbox_storageModuleNotFoundError: No module named 'app'使用 PowerShell 运行 MCP Server 时,请先设置:
$env:PYTHONPATH = "src"
uv run -- python -m app.mcp_server
Prerequisites
Time Estimate
15-60 minutes depending on server complexity
Steps
Troubleshooting
✓ Do
✗ Don't
💡 Pro Tips
Architecture
Model Context Protocol standardizes how AI hosts (Claude, Cursor) communicate with external tools and data sources through server implementations.
Protocols
Compatibility
✓ Use when
Use when you need Claude to access external data, execute actions, or integrate with tools. Best for extending AI capabilities beyond conversation.
✗ Avoid when
Avoid when native integrations exist (use official APIs directly), for real-time critical systems, or when security/compliance requires zero external dependencies.