Are you the author? Sign in to claim
Yes Mcp server in bash
A lightweight, zero-overhead implementation of the Model Context Protocol (MCP) server in pure Bash.
Why? Most MCP servers are just API wrappers with schema conversion. This implementation provides a zero-overhead alternative to Node.js, Python, or other heavy runtimes.
jq for JSON processing (brew install jq on macOS)git clone https://github.com/muthuishere/mcp-server-bash-sdk
cd mcp-server-bash-sdk
chmod +x mcpserver_core.sh moviemcpserver.sh
echo '{"jsonrpc": "2.0", "method": "tools/call", "params": {"name": "get_movies"}, "id": 1}' | ./moviemcpserver.sh
┌─────────────┐ ┌────────────────────────┐
│ MCP Host │ │ MCP Server │
│ (AI System) │◄──────► │ (moviemcpserver.sh) │
└─────────────┘ stdio └────────────────────────┘
│
┌───────┴──────────┐
▼ ▼
┌───────────────────┐ ┌───────────────┐
│ Protocol Layer │ │ Business Logic│
│(mcpserver_core.sh)│ │(tool_* funcs) │
└───────────────────┘ └───────────────┘
│ │
▼ ▼
┌─────────────────┐ ┌───────────────┐
│ Configuration │ │ External │
│ (JSON Files) │ │ Services/APIs │
└─────────────────┘ └───────────────┘
When implementing tool functions for the MCP server, follow these guidelines:
tool_ followed by the same name defined in tools_list.json$1 containing JSON argumentsweatherserver.sh)#!/bin/bash
# Weather API implementation
# Override configuration paths BEFORE sourcing the core
MCP_CONFIG_FILE="$(dirname "${BASH_SOURCE[0]}")/assets/weatherserver_config.json"
MCP_TOOLS_LIST_FILE="$(dirname "${BASH_SOURCE[0]}")/assets/weatherserver_tools.json"
MCP_LOG_FILE="$(dirname "${BASH_SOURCE[0]}")/logs/weatherserver.log"
# MCP Server Tool Function Guidelines:
# 1. Name all tool functions with prefix "tool_" followed by the same name defined in tools_list.json
# 2. Function should accept a single parameter "$1" containing JSON arguments
# 3. For successful operations: Echo the expected result and return 0
# 4. For errors: Echo an error message and return 1
# 5. All tool functions are automatically exposed to the MCP server based on tools_list.json
# Source the core MCP server implementation
source "$(dirname "${BASH_SOURCE[0]}")/mcpserver_core.sh"
# Access environment variables
API_KEY="${MCP_API_KEY:-default_key}"
# Tool: Get current weather for a location
# Parameters: Takes a JSON object with location
# Success: Echo JSON result and return 0
# Error: Echo error message and return 1
tool_get_weather() {
local args="$1"
local location=$(echo "$args" | jq -r '.location')
# Parameter validation
if [[ -z "$location" ]]; then
echo "Missing required parameter: location"
return 1
fi
# Call external API
local weather=$(curl -s "https://api.example.com/weather?location=$location&apikey=$API_KEY")
echo "$weather"
return 0
}
# Start the MCP server
run_mcp_server "$@"
assets/weatherserver_tools.json{
"tools": [
{
"name": "get_weather",
"description": "Get current weather for a location",
"inputSchema": {
"type": "object",
"properties": {
"location": {
"type": "string",
"description": "City name or coordinates"
}
},
"required": ["location"]
}
}
]
}
assets/weatherserver_config.json{
"protocolVersion": "2025-03-26",
"serverInfo": {
"name": "WeatherServer",
"version": "1.0.0"
},
"capabilities": {
"tools": {
"listChanged": true
}
},
"instructions": "This server provides weather information."
}
chmod +x weatherserver.sh
"mcp": {
"servers": {
"my-weather-server": {
"type": "stdio",
"command": "/path/to/your/weatherserver.sh",
"args": [],
"env": {
"MCP_API_KEY": "your-api-key"
}
}
}
}
/mcp my-weather-server get weather for New York
For AI assistants and local tool execution, these aren't blocking issues.
This project is licensed under the MIT License - see the LICENSE file for details.
Blog : https://medium.com/@muthuishere/why-i-built-an-mcp-server-sdk-in-shell-yes-bash-6f2192072279
Run Claude Code as an MCP server so any agent can delegate coding tasks to it
Browser automation using accessibility snapshots instead of screenshots
Secure MCP server for MySQL database interaction, queries, and schema management
Run analytics queries on ClickHouse — explore schemas, execute SQL, fetch results