Bug: Initialize settings in configured workflow task executor

[Robban1980]

Summary

Fixes WorkflowTaskExecutor when it is created with an explicit WorkflowTaskExecutorConfiguration.

The configured constructor did not initialize _workerSettings, causing a null-reference failure when the executor starts or polls for tasks.

Changes

• Initialize _workerSettings from the supplied configuration.
• Fall back to worker.WorkerSettings when no explicit configuration is supplied.
• Add regression coverage for both explicit and fallback configuration paths.

Validation

The new tests verify that polling uses the expected worker ID, domain, and requested task count for both configuration paths.

[Robban1980]

This PR’s integration_tests_oss job reads its image tag from the organization Actions variable E2E_TEST_OSS_CONDUCTOR_VERSION.

In the failing run, that variable resolves to empty, so the job exits before its generated docker-compose-oss.yaml exists. Could an org owner set the variable and grant conductor-oss/csharp-sdk access?

The PR does not need a code change to supply the value, but I updated the failure-log guard so the missing-variable error is no longer obscured by the follow-up compose-file message.

[Robban1980]

Any update?

[chrishagglund-ship-it]

We have merged a change into the main branch that will set a default oss conductor image version to use in situations where the org variable cannot be resolved. I think a new commit or needs to be pushed in this branch to get the CI run to use the latest main workflow, because just re-running it doesn't seem to be picking up the upstream fix.

[Robban1980]

@chrishagglund-ship-it i synced my branch with the current main, let's see if this will work.