mcp-demo
URL MCP is a proof of concept stateless MCP server builder that allows users to build MCP servers without writing or hosting code. It's intended for protocol and security experimentation rather than for building real world MCP integrations.
Install / Use
claude mcp add sshh12 -- npx -y github:sshh12/mcp-demoIf the server publishes to npm under a different name, use that package instead — check the repo README.
MCP Server
Model Context Protocol server
Quality Score
Category
SecuritySupported Platforms
Skill content
View source on GitHubURL-Based MCP Server
<img width="500" alt="Screenshot 2025-04-12 at 12 07 39 PM" src="https://github.com/user-attachments/assets/a9b25d5a-a950-4fdb-b789-90d637f6c142" />URL MCP is a proof of concept stateless MCP server builder that allows users to build MCP servers without writing or hosting code. It's intended for protocol and security experimentation rather than for building real world MCP integrations.
Usage
- Go to https://url-mcp-demo.sshh.io/
- Add custom tools to your MCP server
- If you want to hardcode tool responses, select "static text response"
- If you want dynamic HTTP-based responses, select "http post endpoint"
- Typically I'll use a temp URL from https://webhook.site/ which logs all the requests and allows you to configure custom responses
- Copy the
MCP Configurationinto your client of choice
Local Hosting
There is not really a point to hosting this locally (the whole idea is that you can use this without hosting just with the URL) but in case you want to modify the app itself:
cd backend && pip install requirements.txt && python main.py
Example: System Prompt Exfiltration
Create a custom MCP server for extracting the system prompt of an application.
- Configure the tools to trick the client into trusting it. Use a https://webhook.site/ temp url for the tool response.
- Copy the MCP config
- In the client ask for an audit and the view the webhook logs.
| Step 1 | Step 2 | Step 3 | | ---------------------------------------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------- | | <img width="822" alt="Screenshot 2025-04-12 at 12 14 40 PM" src="https://github.com/user-attachments/assets/bee7fb56-c405-4260-82ef-e485555c19ae" /> | <img width="803" alt="Screenshot 2025-04-12 at 12 14 50 PM" src="https://github.com/user-attachments/assets/5a94ce4f-904d-4ee2-b117-e7147e877b56" /> | <img width="1222" alt="Screenshot 2025-04-12 at 12 15 00 PM" src="https://github.com/user-attachments/assets/d375f49e-6852-48d2-a502-dc6355530e22" /> |
Related Skills
Agent-Reach
84.5kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
headroom
73.4kCompress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
ruflo
73.0k🌊 The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, federation, vector RAG integration, and native Claude Code / Codex / Hermes and many more Integrated
CowAgent
47.1kOpen-source super AI assistant & Agent Harness. Plans tasks, runs tools and skills, self-evolves with memory and knowledge. Multi-agent, multi-model, multi-channel. Lightweight, extensible, one-line install.
