EdgeKit: Cloudflare Workers Python native SDK and builder toolkit

EdgeKit has two jobs:

1. Provide a small Python-first API for writing Workers Python.
2. Build a runtime bundle by analyzing your project, tree-shaking, vendoring dependencies.

What EdgeKit Provides

• WorkerEntrypoint[Env] for typed Worker classes
• Request, Response, Headers, and URL wrappers around the Workers Web APIs
• Typed binding wrappers for:
  • static assets
  • D1
  • KV
  • R2
  • Queues
  • Durable Objects
• WSGI and ASGI adapters for framework integration
• A CLI for:
  • project analysis
  • risk checks for tree-shaking and bundling
  • bundle emission
  • report rendering

Requirements

• Python >=3.12
• uv for dependency management
• wrangler CLI for development and deployment

Local Development

Install Python and Node dependencies in the repository root:

uv sync

Useful commands:

edgekit analyze
edgekit doctor
edgekit build
npm run dev

The root wrangler.jsonc should be configured to call edgekit build before wrangler dev.

Quick Start

Initialize a new project first:

uv run pywrangler init
cd <your-project>
uv add edgekit

The minimal Worker shape looks like this:

from edgekit import Request, Response, WorkerEntrypoint


class Default(WorkerEntrypoint):
    async def fetch(self, request: Request) -> Response:
        return Response.json(
            {
                "ok": True,
                "method": request.method,
                "pathname": request.url.pathname,
            }
        )

Set the Worker entrypoint in pyproject.toml:

[tool.edgekit.builder]
entry = "src/app.py"
compatibility_date = "2026-04-13"

Then point Wrangler at the built output:

{
  "main": "build/edgekit/wrangler/python_modules/app.py",
  "compatibility_date": "2026-04-13",
  "compatibility_flags": ["python_workers"],
  "build": {
    "command": "uv run edgekit build",
  },
}

Run:

wrangler dev

Typed Environment Bindings

EdgeKit can bind the Workers env object to a typed Python protocol or class.

from typing import Protocol

from edgekit import Request, Response, WorkerEntrypoint
from edgekit.bindings import D1Database, StaticAssets


class Env(Protocol):
    ASSETS: StaticAssets
    DB: D1Database


class Default(WorkerEntrypoint[Env]):
    async def fetch(self, request: Request) -> Response:
        ok = await self.env.DB.prepare("select 1 as ok").first("ok", type=int)
        asset = await self.env.ASSETS.fetch("/hello.txt")
        return Response.json(
            {
                "ok": ok if ok is not None else 0,
                "asset": await asset.read_text(),
            }
        )

Runtime Helpers

edgekit.runtime exports two helpers:

• current_env([EnvType])
  • Returns the active env object for the current request scope.
  • Used by adapters such as WSGI and ASGI to expose the current Worker env inside framework code.
• await_sync(awaitable)
  • Runs an awaitable synchronously via Pyodide runtime bindings.

Example:

from typing import Protocol

from edgekit.adapters import WSGI
from edgekit.bindings import StaticAssets
from edgekit.runtime import await_sync, current_env


class Env(Protocol):
    ASSETS: StaticAssets


def read_asset_text(path: str) -> str:
    env = current_env(Env)
    response = await_sync(env.ASSETS.fetch(path))
    return await_sync(response.read_text())

License

MIT License