cloudflare

For CloudFlare development, deployment, e.g. Python CloudFlare Workers

$ 설치

git clone https://github.com/sanand0/scripts /tmp/scripts && cp -r /tmp/scripts/agents/cloudflare ~/.claude/skills/scripts

// tip: Run this command in your terminal to install the skill


name: cloudflare description: For CloudFlare development, deployment, e.g. Python CloudFlare Workers

General

  • Modifying code while wrangler dev is running slows it down. Stop, edit, then restart.

Python Cloudflare Workers

Sample wrangler.toml:

name = "..."
main = "..."
compatibility_date = "..."
compatibility_flags = ["python_workers"]

[observability]
enabled = true

Sample entry.py:

from workers import Response, WorkerEntrypoint

class Default(WorkerEntrypoint):
    async def fetch(self, request, env) -> Response:
        return Response('{"x": 1}', headers={"content-type": "application/json"})

To integrate a FastAPI app:

from workers import WorkerEntrypoint
from my_fastapi_app import app
import asgi

class Default(WorkerEntrypoint):
   async def fetch(self, request):
       return await asgi.fetch(app, request.js_object, self.env)

Usage notes:

  • Ensure there's a pyproject.toml with dependencies.
  • Prep env: uv add --dev workers-py workers-runtime-sdk
  • Vendor deps (if any): uv run pywrangler sync (uses python_modules).
  • Dev: uv run pywrangler dev then hit http://localhost:8787.
  • Deploy: uv run pywrangler deploy.
  • Verify: curl https://<worker-name>.<account>.workers.dev.