File size: 6,382 Bytes
5233b57 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 | """
dispatchAI MCP Server - gives the agent tools to interact with the phone farm API.
This is a simple MCP server that exposes the dispatchAI API as tools.
"""
import json
import httpx
import asyncio
from typing import Any
# MCP server using the simple JSON-RPC protocol
# This runs as a subprocess that Hermes communicates with via stdio
BASE_URL = "http://127.0.0.1:8081"
API_KEY = "da-demo-key-0001"
TOOLS = [
{
"name": "chat_completion",
"description": "Send a chat completion request to a dispatchAI model running on a phone. Returns the model's response.",
"inputSchema": {
"type": "object",
"properties": {
"message": {"type": "string", "description": "The user message to send"},
"model": {"type": "string", "description": "Model name (default: SmolLM2-135M)", "default": "dispatchAI/SmolLM2-135M-Instruct-mobile"},
"max_tokens": {"type": "integer", "description": "Max tokens to generate", "default": 100},
},
"required": ["message"],
},
},
{
"name": "list_models",
"description": "List all available dispatchAI models",
"inputSchema": {"type": "object", "properties": {}},
},
{
"name": "get_phone_status",
"description": "Get the status of the phone farm - how many phones are connected and their load",
"inputSchema": {"type": "object", "properties": {}},
},
{
"name": "get_api_usage",
"description": "Get API usage statistics for the current API key",
"inputSchema": {"type": "object", "properties": {}},
},
]
async def call_tool(name: str, arguments: dict) -> str:
"""Call a tool and return the result."""
async with httpx.AsyncClient(timeout=120.0) as client:
if name == "chat_completion":
model = arguments.get("model", "dispatchAI/SmolLM2-135M-Instruct-mobile")
message = arguments.get("message", "")
max_tokens = arguments.get("max_tokens", 100)
r = await client.post(
f"{BASE_URL}/v1/chat/completions",
headers={
"Content-Type": "application/json",
"Authorization": f"Bearer {API_KEY}",
},
json={
"model": model,
"messages": [{"role": "user", "content": message}],
"max_tokens": max_tokens,
},
)
if r.status_code == 200:
data = r.json()
content = data.get("choices", [{}])[0].get("message", {}).get("content", "")
phone = data.get("phone_info", {})
usage = data.get("usage", {})
return json.dumps({
"response": content,
"phone": phone.get("serial", "unknown"),
"speed_tps": phone.get("generation_tps", 0),
"tokens_used": usage.get("total_tokens", 0),
})
else:
return f"Error: {r.status_code} - {r.text[:200]}"
elif name == "list_models":
r = await client.get(
f"{BASE_URL}/v1/models",
headers={"Authorization": f"Bearer {API_KEY}"},
)
if r.status_code == 200:
data = r.json()
models = [m["id"] for m in data.get("data", [])]
return json.dumps({"models": models})
return f"Error: {r.status_code}"
elif name == "get_phone_status":
r = await client.get(
f"{BASE_URL}/admin/phones",
headers={"Authorization": f"Bearer {API_KEY}"},
)
if r.status_code == 200:
return json.dumps(r.json())
return f"Error: {r.status_code}"
elif name == "get_api_usage":
r = await client.get(
f"{BASE_URL}/v1/usage",
headers={"Authorization": f"Bearer {API_KEY}"},
)
if r.status_code == 200:
return json.dumps(r.json())
return f"Error: {r.status_code}"
return f"Unknown tool: {name}"
# Simple MCP JSON-RPC server over stdio
import sys
def main():
"""Main MCP server loop - reads JSON-RPC from stdin, writes to stdout."""
while True:
try:
line = sys.stdin.readline()
if not line:
break
req = json.loads(line)
method = req.get("method", "")
req_id = req.get("id")
if method == "initialize":
resp = {
"jsonrpc": "2.0",
"id": req_id,
"result": {
"protocolVersion": "2024-11-05",
"capabilities": {"tools": {}},
"serverInfo": {"name": "dispatchai-mcp", "version": "1.0.0"},
},
}
sys.stdout.write(json.dumps(resp) + "\n")
sys.stdout.flush()
elif method == "tools/list":
resp = {
"jsonrpc": "2.0",
"id": req_id,
"result": {"tools": TOOLS},
}
sys.stdout.write(json.dumps(resp) + "\n")
sys.stdout.flush()
elif method == "tools/call":
params = req.get("params", {})
tool_name = params.get("name", "")
arguments = params.get("arguments", {})
result = asyncio.run(call_tool(tool_name, arguments))
resp = {
"jsonrpc": "2.0",
"id": req_id,
"result": {
"content": [{"type": "text", "text": result}]
},
}
sys.stdout.write(json.dumps(resp) + "\n")
sys.stdout.flush()
except Exception as e:
if req_id:
resp = {
"jsonrpc": "2.0",
"id": req_id,
"error": {"code": -32603, "message": str(e)},
}
sys.stdout.write(json.dumps(resp) + "\n")
sys.stdout.flush()
if __name__ == "__main__":
main()
|