The current way of integrating MCP in stdio mode is to wrap the stdio into a SSE listener. It requires another pod to run. If the original MCP supports streamHTTP, then the wrapper is really NOT necessary. Can mcp-context-forge allow adding an MCP server running in the same container as a subprocess using stdio? just like configuring a stdio MCP in cursor as follows (https://github.com/instana/mcp-instana?tab=readme-ov-file#stdio-mode-1):
{
"servers": {
"Instana MCP Server": {
"command": "mcp-instana",
"args": ["--transport", "stdio"],
"env": {
"INSTANA_BASE_URL": "https://your-instana-instance.instana.io",
"INSTANA_API_TOKEN": "your_instana_api_token"
}
}
}
}
🧭 Type of Feature
Please select the most appropriate category:
┌─────────────────────────────────────────────────────────────┐
│ Local Machine │
│ │
│ ┌────────────────────────────────────────────────────┐ │
│ │ MCP Context Forge Gateway (Python) │ │
│ │ - Runs locally (not in Docker) │ │
│ │ - Port: 4444 │ │
│ │ - Manages MCP server connections │ │
│ │ │ │
│ │ ┌──────────────────────────────────────┐ │ │
│ │ │ mcp-instana (Python) │ │ │
│ │ │ - Runs from local folder via uv │ │ │
│ │ │ - Embedded via stdio transport │ │ │
│ │ │ - Connects to remote Instana API │ │ │
│ │ └──────────────────────────────────────┘ │ │
│ └────────────────────────────────────────────────────┘ │
│ │ │
│ │ connects to │
│ ▼ │
│ ┌────────────────────────────────────────────────────┐ │
│ │ Docker Containers │ │
│ │ ┌──────────────┐ ┌──────────────┐ │ │
│ │ │ PostgreSQL │ │ Redis │ │ │
│ │ │ Port: 5432 │ │ Port: 6379 │ │ │
│ │ └──────────────┘ └──────────────┘ │ │
│ │ │ │
│ │ ┌──────────────┐ ┌──────────────┐ │ │
│ │ │ pgAdmin │ │ Redis Insight│ │ │
│ │ │ Port: 5050 │ │ Port: 5540 │ │ │
│ │ └──────────────┘ └──────────────┘ │ │
│ └────────────────────────────────────────────────────┘ │
└─────────────────────────────────────────────────────────────┘
│
│ API calls
▼
┌─────────────────────────┐
│ Remote Instana API │
│ (your-tenant.instana.io)│
└─────────────────────────┘
The current way of integrating MCP in stdio mode is to wrap the stdio into a SSE listener. It requires another pod to run. If the original MCP supports streamHTTP, then the wrapper is really NOT necessary. Can mcp-context-forge allow adding an MCP server running in the same container as a subprocess using stdio? just like configuring a stdio MCP in cursor as follows (https://github.com/instana/mcp-instana?tab=readme-ov-file#stdio-mode-1):
🧭 Type of Feature
Please select the most appropriate category: