Allow fallback to local bindings server if remote is down

[krzema12]

Problem

We've recently had an incident where the bindings server under https://bindings.krzeminski.it/ was down or unstable. Because of this, the users experienced CI failures and couldn't develop the workflows.

Proposed approach

Let's add a mode where the logic detects that the remote server is down, and a local instance of the server is started. To start with, it should be an opt-in feature, but in the longer run, we can make it the default behavior.

Let's build a PoC first: #1766

[krzema12]

Demo of a working PoC: https://github.com/typesafegithub/github-workflows-kt/actions/runs/12723179689/job/35468014941?pr=1766

[krzema12]

Here's a feature that would let the users mitigate the problem only for the consistency check jobs (running on CI): #1767

[krzema12]

Leaving open to remember about at least trying to cover other cases, like:

• local development - recommend starting the service locally
• CI checks that run consistency check for all existing workflows, also e.g. schedule-based