[adewale]

What and why

Adopt Ruff 0.16.0 and restore the existing Linux Workers integration suite. The failing integrations also fail without the Ruff changes: dependency and Workers runtime drift prevented the original examples from starting.

Changes

• Pin Ruff and ruff-pre-commit to 0.16.0, apply its default-rule fixes, and run Ruff in CI.
• Remove typing-only webtypy from runtime dependencies. Pin workers-py to 1.17.7 and CI's uv to 0.12.3; explicitly use Python 3.13.
• Import the FastAPI app inside the request handler. Current FastAPI telemetry needs randomness unavailable during Worker startup; routes and responses are unchanged.
• Remove D1's unused AI binding, which otherwise requests Cloudflare credentials in local development.
• Enable python_workflows, now required for the Python workflow API, and verify the DAG reaches complete rather than merely returning JSON.
• Use isolated temporary Worker storage (including D1 initialization) and terminate the entire dev-server process group between tests.
• Allow cold dependency downloads for every example in CI and fail promptly if a server exits before becoming ready.

Validation

• Linux Docker, Python 3.13, workers-py 1.17.7, Wrangler 4.149.0: uv run pytest -vv -s — 8 passed, 1 xfailed, 1 xpassed.
• uv run ruff check . and staged diff whitespace checks pass.
• Before the fixes, the integration suite reproduced Worker setup/HTTP failures; after the fixes, FastAPI, D1, Durable Objects, Cron and a complete Workflow all pass.
• The existing Workers AI expected failure remains. The LangChain XPASS is an existing placeholder test, not proof of a working LangChain integration. No production deployment was performed.

Risk

Moderate: this updates the Worker development toolchain across the examples. The live HTTP integrations exercise the changed runtime paths; GitHub's clean Linux CI run is the final merge gate.