floatdrop / moq-go / 31660316794
89%

DEFAULT BRANCH: draft-19
Ran
Jobs 1
Files 100
Run time 1min
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
coverage: 85.827%. First build
31660316794

push

github

web-flow 
ci: publish coverage to Coveralls instead of a hand-refreshed badge (#66)

The README badge was a percentage baked into a shields.io URL, rewritten only
when someone ran `make cover-badge`. A number that moves only when a human
moves it drifts, and it already had: 86% committed against 87.7% actual
after the tests in #64 and #65.

CI now uploads the coverage profile on every build and Coveralls owns the
badge. The repo is public, so the action authenticates with the built-in
GITHUB_TOKEN — no repo secret, nothing to configure for a fork.

What is published is deliberately not what is gated, and the workflow says
so: Coveralls gets the -coverpkg whole-suite profile over ./pkg/..., the same
measurement `make cover-total` reports, while coverage-floors.txt keeps the
per-package floors that credit a package only for what its own tests reach.
The floors stay the gate because that is what makes a regression
attributable; Coveralls gates nothing. cmd/* and the etcd/nats submodules are
outside the published profile, stated in the workflow and README rather than
left to be discovered.

Also deletes what this makes dead — the --badge mode of
scripts/coverage-total.sh, its README-rewriting sed, and the make cover-badge
target — and stops CLAUDE.md quoting a figure that will rot.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>

8775 of 10224 relevant lines covered (85.83%)

1.93 hits per line

Jobs
ID Job ID Ran Files Coverage
1 31660316794.1 100
85.83
 GitHub Action Run
Source Files on build 31660316794