[tyrsson]

Why

coverage.status.patch.default.target was auto. Codecov resolves auto from the base commit's coverage report. The org CI workflow at webinertia/.github, .github/workflows/continuous-integration.yml, runs on pull_request only, so a default branch can have no coverage report at all.

That leaves the patch gate in one of two bad states:

• With no base report, Codecov posts a passing status titled "No report found to compare against", which gates nothing.
• With a stale base report, the gate compares the diff against an arbitrary old snapshot.

Either way the gate is not measuring the diff.

Change

target: auto becomes target: 95%, with a comment recording why. 95 matches the fleet mutation gate (min_msi and min_covered_msi are both 95 in webware-ci.json).

threshold: 0% and base: auto are unchanged. coverage.status.project.default stays auto: a single shared number cannot serve a fleet whose project coverage varies widely.

Precedent

Already proven in webinertia/webware-tools#70 (merged as 4c3dbf0) and webinertia/asciidoc-php#19 (merged as 32bd9f2). After that fix, asciidoc-php-multi-repo#8 head 7ea871f reports codecov/patch as "97.44% of diff hit (target 95.00%)" instead of "No report found to compare against".

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!