Repository navigation
chore: release v2.8.0 — every workflow runs on GitHub-hosted runners - #70
Conversation
protoLabs no longer uses Namespace runners for anything. - ci.yml, workflow-security-lint.yml, tauri-release.yml (publish job): runs-on: ubuntu-latest. The tauri build matrix was already hosted. - plugin-release.yml: the `runner` input stays, defaults to ubuntu-latest; vars.NSC_RUNNER is gone. - workspace-config: the rule is inverted and renamed workflows-use-owned-runners -> workflows-use-github-hosted-runners. A namespace-profile-*/nscloud-* label or any NSC_RUNNER reference is now the error; the fix is ubuntu-latest (or the matching hosted label). The `# workspace-config: allow-hosted-runner` comment is no longer read (it is harmless if left in). listRunnerExceptions() and --json runnerExceptions are kept for compatibility and are always empty. - Docs and tests updated to match. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
There was a problem hiding this comment.
QA panel review — WARN
code-review-structural · head e9cad937a939 · formal
Low-risk rule inversion with one confirmed minor: the README (line 292) and docs/index.md (line 34) still describe the old self-hosted-runner standard after the rule was flipped to workflows-use-github-hosted-runners. Fix the stale prose before merge. Verification confirmed the finding at head SHA; no panel disagreement, no gaps.
Findings
| Severity | Location | Finding | Verified | |
|---|---|---|---|---|
| 🟡 | minor | README.md:292 |
README still says the workspace-config standard has a self-hosted-runner check that stops repos burning GitHub-hosted minutes. This PR inverted that rule, so t… | confirmed |
findings JSON (machine-readable)
[
{
"file": "README.md",
"line": 292,
"severity": "minor",
"category": "removed-behavior",
"claim": "README still says the workspace-config standard has a self-hosted-runner check that stops repos burning GitHub-hosted minutes. This PR inverted that rule, so the text now describes the opposite of what the check does.",
"evidence": "README.md:292-293: \"fleet-watched repo should carry \u2014 plus a self-hosted-runner check so no repo\" / \"silently burns GitHub-hosted minutes.\" The rule is now `workflows-use-github-hosted-runners`, which flags Namespace runners and treats GitHub-hosted runners as the standard. `docs/index.md:34` also still says \"why owned runners\" for the page this PR rewrote as \"GitHub-hosted runners, not Namespace\".",
"verdict": "confirmed",
"note": "Read README.md:292-293 at head SHA: text verbatim matches the finding's quote. Read docs/index.md:34: still says \"why owned runners\". PR description confirms the rule was inverted to workflows-use-github-hosted-runners, so both passages are stale."
}
]There was a problem hiding this comment.
Promoting the WARN verdict for head e9cad937a939: all checks terminal-green, zero unresolved review threads. (approve-on-green)
Open findings carried by this approval — non-blocking, but they did not go away:
- minor
README.md:292— README still says the workspace-config standard has a self-hosted-runner check that stops repos burning GitHub-hosted minutes. This PR inverted that rule, so the text now describes the opposite of wha
Approving a WARN does not resolve its findings (issue #22).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
There was a problem hiding this comment.
QA panel review — PASS
code-review-structural · head ecdc14f7603f · formal
No findings were raised, so there was nothing to verify. The panel's synthesized result was an explicit empty array and the verifier confirmed nothing-to-verify. The round-1 minor (README.md:292 stale wording) is resolved by this PR's wording fix; no blocker or major priors remain open. No coverage gaps to report.
No findings — the review came back clean.
findings JSON (machine-readable)
[]Incremental re-review (#313): 2 of 18 file(s) changed since head e9cad937a939 (or depend on a change) and were reviewed. 16 unchanged file(s) were not re-read. 0 finding(s) on them are carried forward unchanged.
Comment @vera review for a full re-review.
Addressed since the last round (1 minor/nit): README.md:292 — no longer raised, and the PR changed the code around them since they were raised.
protoLabs no longer uses Namespace runners for anything. This makes release-tools match, and releases it so
@v2consumers get it.Changes
ci.yml,workflow-security-lint.ymland thepublishjob intauri-release.ymlnow useruns-on: ubuntu-latest. The Tauri build matrix already ran on hosted runners and is unchanged.plugin-release.yml(the reusable workflow every plugin calls at@v2): therunnerinput is kept, its default is nowubuntu-latest, andvars.NSC_RUNNERis no longer read.workflows-use-owned-runnersbecomesworkflows-use-github-hosted-runners. GitHub-hosted labels (ubuntu-*,macos-*,windows-*) are the standard. Anamespace-profile-*ornscloud-*label, or anyNSC_RUNNERreference (including${{ vars.NSC_RUNNER || '...' }}expressions), is now the error. The fix it suggests isubuntu-latestor the matching hosted label.# workspace-config: allow-hosted-runnercomment: no longer read. Hosted runners need no exception now, so the comment does nothing. Repos that still have it pass, and it can be deleted whenever someone next edits the file.listRunnerExceptions()and the--jsonrunnerExceptionskey are kept so nothing that imports or parses them breaks. Both are always empty now.Version is bumped to 2.8.0. Merging this
chore: releasecommit runsrelease.yml, which tags v2.8.0 and movesv2.Verification
npm test: 93/93 passing.npm run lintandnpm run smokeare clean.verify-workspace-configon this repo reports no runner violation.🤖 Generated with Claude Code