docs: say in the README that this is a fork, and what that changes - #6
Open
theRealBithive wants to merge 4 commits into
Open
theRealBithive wants to merge 4 commits into
theRealBithive wants to merge 4 commits into
Conversation
The README was upstream's product pitch, and it still is below the new section — deliberately, so an upstream sync stays a sync rather than a merge argument. What it never said is that this repository is a fork Exkulpa runs as a network service, which is the fact AGPL-3.0 §13 attaches obligations to, and that four fork-local rules apply that upstream has no reason to carry. The badges are rebuilt around what is true here. The CI badge now points at this fork's workflow instead of upstream's, so it reports our runs and not somebody else's. The clone commands point at this fork for the same reason. The two gate badges deliberately state a rule rather than a number. There is no repository-wide coverage or mutation percentage to report: both gates grade what a change touches, and a repo total is a figure nobody has to defend which improves when you measure less. A percentage badge would have had to be invented or fetched from a third party, and the diff-scoped mutation score in particular covers five declared files — printing it as a repository score would be a straightforward misrepresentation. Since both gates are jobs inside the CI workflow, the CI badge already goes red when either fails, which is the honest live signal. The stars badge is gone. On a fork nobody stars it measured nothing. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
…costs Adding a manifest entry asserts that the named suites pin the file. On a pre-existing file that claim is usually false, and the only way to find out is to run the gate. One file this round cost a dozen small tests to make the claim true; the next would have cost sixty mutants of untouched error handling and was backed out instead. Also records two traps that each cost a run: NoCoverage mutants mean the entry names too few suites rather than that the code is untested, and editing the manifest as text puts the new record in `equivalents` too, which the gate reports as a missing reason. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The badge blockquote claimed the CI badge goes red when `diff-coverage` or `mutation` fails. True as far as it goes — neither job is `continue-on-error` and neither is skipped on a push — but it reads as if the green badge on `main` were evidence those two gates found something. It is not: both compare against the merge base, and on `main` that is `main`, so the diff is empty and they grade nothing. The evidence is on the pull request, which is what the badge label says and what the sentence now says too. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What changes
The README was upstream's, unchanged, pointing at upstream's repository. Anyone
landing here cloned the wrong remote and read CI badges for a workflow that is
not the one running.
theRealBithive/quackback, twopolicy badges, the upstream link, the licence. The stars badge is gone — it
counted upstream's.
## This is a forksection at the top. Everything below it is upstream'sown description of the product and is left exactly as it stands, so a sync
stays a sync rather than a merge.
contributors list still point upstream on purpose: those are upstream's and we
do not host them.
Why there is no coverage percentage
There was a live number to be had and it would have been a misrepresentation.
This repository has no repository-wide coverage or mutation percentage,
deliberately: both gates grade what a change touches, not the total. A total
is a number nobody has to defend, and it improves when you measure less — which
is the exact failure mode
coverage-scope.test.tsandmutation-scope.test.tsexist to make loud.
So the two blue badges state a rule and say so in the text under them, and the
one live badge is CI. The blockquote also says where the two gates actually
bite: they compare against the merge base, so on
mainthe diff is empty andthey grade nothing. The evidence is on the pull request.
Both badge URLs were fetched and render (
CI: passing,diff coverage: enforced per PR), and bothCLAUDE.mdanchors resolve.Also in here
An entry in
SELF-IMPROVE.md: declaring a file in the mutation manifest is anassertion that the named suites pin it on their own, and finding out that the
assertion is false costs a full run. That happened twice on the branch this was
written alongside.
Docs only. Nothing in the running service changes.
🤖 Generated with Claude Code