sahilTiwariiiii/mcp_server
3.2
If you are the rightful owner of mcp_server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.
This document provides a structured summary of the Model Context Protocol (MCP) server for the Statistics Canada LODE Dataset, detailing its features, tools, resources, usage, and FAQs.
Tools
4
Resources
0
Prompts
0
MCP Server for Statistics Canada LODE Dataset (HTTP/SSE)
Overview
- Public, no-auth Model Context Protocol (MCP) server over HTTP + SSE.
- Exposes compact tools to browse schema and query data with safe limits.
- Tested locally in VS Code MCP chat; deployable to EC2 for public access.
Links
- Docker image: https://hub.docker.com/r/samrat156/mcpserver
- GitHub repo: https://github.com/sahilTiwariiiii/mcp_server
- Live EC2 (SSE): [http://15.206.164.94:5000/sse]
- Companion POST: [http://15.206.164.94:5000/sse]
- Testing media (video/photos): [https://drive.google.com/drive/folders/1zBW7JeyspttDaB0jnbvPWusZDf-xptEo?usp=sharing]
Endpoints
- SSE stream:
GET /sse - Companion POST:
POST /sse/messages(JSON-RPC)
Tools
describe_dataset– Summarize fields, types, units, and source links.query_data– Filter byprovince,city,sector,naics; supportslimit≤ 1000.get_metadata– Release info, licence, providers sample, official links.get_download_links– Official catalogue page and local dataset file paths.
Quick Start (Docker)
- Pull:
docker pull samrat156/mcpserver:latest - Run:
docker run --name mcpserver -d -p 3100:3100 -e PORT=3100 samrat156/mcpserver:latest - Test SSE: open
http://localhost:3100/sse - Send a request:
POST http://localhost:3100/sse/messageswith JSON-RPC (e.g.,tools/list).
VS Code Integration (MCP)
- Create
.vscode/mcp.jsonentry: { "servers": [{ "name": "lode-odbus-mcp", "type": "sse", "url": "http://localhost:3100/sse", "messagesUrl": "http://localhost:3100/sse/messages" }] }
Challenge Notes (48-Hour: “MCP for a Statistics Canada LODE Dataset”)
- Transport: HTTP + SSE; anonymous, no auth.
- Tooling: discover (
describe_dataset), query (query_data), helpers (get_metadata,get_download_links). - Inputs validated; friendly errors for over-broad filters; query limits enforced.
- Latency: typical returns in ~5–10s; warmed queries faster.
- Minimal footprint: no tracking, no analytics, local logs only.
Dataset & Licence
- Uses open data with attribution to the Open Government Licence – Canada.
- Licence: https://open.canada.ca/en/open-government-licence-canada
- Metadata and responses include sources and attribution where applicable.
Architecture (brief)
- Node/TypeScript server; SSE stream for events + companion POST for JSON-RPC.
- Local CSVs bundled; loaded at startup; normalized JSON returned.
- Defensive rate limits and input validation; schema surfaced for tooling.
EC2 Deployment (example)
- Build image or pull:
docker pull samrat156/mcpserver:latest - Run on EC2:
docker run -d -p 3100:3100 -e PORT=3100 samrat156/mcpserver:latest - Provide public SSE URL in clients:
https://<ec2-host-or-ip>:3100/sse
Notes
- Large files tracked via Git LFS; collaborators should run
git lfs installonce. - Replace placeholders above (EC2 host/IP and Google Drive link) with your actual URLs.