{"id":46966000,"title":"Show HN: PolyMCP – AI-Callable Python and TS Tools with Inspector and Apps","points":1,"user":"justvugg","time":1770753823,"time_ago":"8 hours ago","type":"link","content":"

I built PolyMCP, an open-source framework around the Model Context Protocol (MCP) that lets you turn any existing Python function into an MCP tool usable by AI agents — with no rewrites, no glue code, no custom wrappers.

Over the last weeks, PolyMCP has grown into a small ecosystem:\n • PolyMCP (core) – expose Python functions as MCP tools\n • PolyMCP Inspector – visual UI to explore, test, and debug MCP servers\n • PolyMCP MCP SDK Apps – build MCP-powered apps with tools + UI resources

⸻

1) Turn any Python function into an MCP tool

Basic example:

from polymcp import expose_tools_http

def add(a: int, b: int) -> int:\n """Add two numbers"""\n return a + b

app = expose_tools_http(\n tools=[add],\n title="Math Tools"\n)

Run it:

uvicorn server_mcp:app --reload

Now add is an MCP-compliant tool that any AI agent can discover and call.

No decorators, no schema files, no agent-specific SDKs.

⸻

2) Real APIs, not toy examples

Existing API code works as-is:

import requests\nfrom polymcp import expose_tools_http

def get_weather(city: str):\n """Return current weather data for a city"""\n response = requests.get(\n f"https://api.weatherapi.com/v1/current.json?q={city}"\n )\n return response.json()

app = expose_tools_http([get_weather], title="Weather Tools")

Agents can now call:

get_weather("London")

and receive real-time data.

⸻

3) Business & internal workflows

Example: internal reporting logic reused directly by agents.

import pandas as pd\nfrom polymcp import expose_tools_http

def calculate_commissions(sales_data: list[dict]):\n """Calculate sales commissions from sales data"""\n df = pd.DataFrame(sales_data)\n df["commission"] = df["sales_amount"] * 0.05\n return df.to_dict(orient="records")

app = expose_tools_http([calculate_commissions], title="Business Tools")

No rewriting legacy logic.

4) PolyMCP Inspector (visual debugging)

To make MCP development usable in practice, I added PolyMCP Inspector:\n • Visual UI to browse tools, prompts, and resources\n • Call MCP tools interactively\n • Inspect schemas, inputs, outputs, and errors\n • Multi-server support (HTTP + stdio)\n • Built-in chat playground (OpenAI / Anthropic / Ollama)

Think “Postman + DevTools” for MCP servers.

Repo: https://github.com/poly-mcp/PolyMCP-Inspector

⸻

5) MCP SDK Apps (tools + UI)

The latest addition is PolyMCP MCP SDK Apps:\n • Build MCP apps, not just tools\n • Expose:\n • tools\n • UI resources (HTML/JS dashboards)\n • app-level workflows\n • Let agents interact with both tools and UIs

This is useful for:\n • internal copilots\n • ops dashboards\n • support tools\n • enterprise AI frontends

Repo: https://github.com/poly-mcp/PolyMCP-MCP-SDK-Apps

⸻

Why this matters (especially for companies)\n • Reuse existing code immediately (scripts, APIs, internal libs)\n • Standard MCP interface instead of vendor-specific agent SDKs\n • Multiple tools, one server\n • Agent-driven orchestration, not hardcoded flows\n • Faster AI adoption without refactoring everything

PolyMCP treats AI agents as clients of your software, not magic wrappers around it.

⸻

Repos\n • Core framework: https://github.com/poly-mcp/PolyMCP\n • Inspector UI: https://github.com/poly-mcp/PolyMCP-Inspector\n • MCP SDK Apps: https://github.com/poly-mcp/PolyMCP-MCP-SDK-Apps

Happy to hear feedback from people building MCP servers, agents, or internal AI tools.","url":"item?id=46966000","comments":[],"comments_count":0}