test: PR size label workflow validation (demo data, do not merge)

[forge-orion]

Test PR — do not merge

This is a test pull request with dummy data only, opened to verify the recently merged PR size label workflow (.github/workflows/pr-size.yml).

This PR was opened by the AI assistant Orion (GitHub user @forge-orion) acting on behalf of @weselben, who owns the fork organization weselben-orion. It will be closed again once the size label has been observed.

Composition

File	Lines	Bucket in workflow
internal/providers/demosize/provider_test.go	200	excluded (*_test.go pathspec)
internal/demosize/doc.go	20	counted (non-test)
docs/demosize/pr-size-label-demo.md	280	counted (non-test)

Effective changed lines expected by the workflow: 300 non-test + 200 excluded test lines → label should be size:L (100–499) with "test lines excluded" classification.

All content is placeholder demo data; nothing here compiles into or affects the gateway (the test file is behind //go:build ignore, the Go package is an unused stub, the markdown is filler).

[coderabbitai]

Important

Draft PR not reviewed

Draft PRs are not automatically reviewed by default.

• Trigger a manual review

To automatically review draft PRs, update your CodeRabbit configuration:

reviews:
  auto_review:
    drafts: true

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[forge-orion]

Post-test analysis — PR size label workflow ✔

This test PR was opened by the AI assistant Orion (@forge-orion) acting on behalf of @weselben. Workflow validation complete; findings below.

Composition (all demo data, zero production impact)

File	Lines	Expected treatment
internal/providers/demosize/provider_test.go	200	excluded via *_test.go pathspec
internal/demosize/doc.go	20	counted
docs/demosize/pr-size-label-demo.md	280	counted

Expected vs actual

Metric	Expected	Actual (workflow log)
Total changed lines	500	— (not directly logged)
Non-test lines	300	300 ✔
Excluded test lines	200	200 ✔
Effective lines	300	300 ✔
Label	size:L (100–499)	size:L ✔
Classification	test lines excluded	test lines excluded ✔

Workflow log line (run 39199161343):

PR #1207: 300 non-test lines, 200 test lines, 300 effective lines -> size:L (test lines excluded)

Observations

• Mixed-PR exclusion works as designed: 40% of the raw diff (200/500 lines) was test-file churn, correctly dropped before banding. Without the exclusion this would have landed at the top of size:L (500 raw lines is the XL boundary) — with it, a mid-band L.
• pull_request_target fired on opened from a fork PR and completed in ~17s. No permission or cross-fork ref fetch issues; refs/pull/1207/head resolved cleanly.
• Label applied on a draft PR as expected (workflow triggers include converted_to_draft/ready_for_review, but opened already covers draft creation).
• Locally reproduced the numstat math against upstream/main...HEAD with the same pathspec set — numbers match the workflow exactly, so the exclusion pathspecs behave identically on a plain git client.

No anomalies found. Closing this test PR; branch test/pr-size-label-demo on weselben-orion/GoModel can be deleted.