[afonsobspinto]

Closes https://metacell.atlassian.net/browse/CH-300

Upgrade Argo Workflows from 3.1.15 to 3.3.10 to enable native self-reporting progress

Changes

• Update controller, server, executor and the seven official 3.3.10 CRDs together.
• Retain the existing PNS configuration and default pod naming to v1 for CloudHarness logs. Existing extraEnv entries, including explicit pod-name overrides, remain intact.
• Remove the unsupported pod-worker argument
• Grant the controller the upstream TaskSet management and TaskResult watch/cleanup permissions.
• Keep lightweight Helm/chart regression checks.

Misc:

• Upgrade Ubuntu version on the scanner-runner
• Update and fix the pre-existing syntax/formatting failures required for CI.

How to test

1. Deploy this branch to a test environment through the usual CloudHarness deployment process.
2. Run an existing workflow through CloudHarness and confirm its result and logs are unchanged.
3. Run a container/script task with the progress configuration below. Confirm intermediate progress while it is running and the final completed count.

Self-reporting progress

Self-reporting is optional; existing tasks need no changes. If enabled, the file format must be completed/total: for example, 2/4 means two of four batches have finished.

The task overwrites $ARGO_PROGRESS_FILE as work advances:

echo "0/4" > "$ARGO_PROGRESS_FILE"
# Process the first batch.
echo "1/4" > "$ARGO_PROGRESS_FILE"
# Continue updating after each batch, ending with "4/4".

Progress is available through the existing CloudHarness workflow API. Reporting is polled, so leave time between updates to observe intermediate values.

No files under libraries/ change.