Skip to main content
KetQat

Appearance

Open-source metrics

Adoption metrics

How much this project is actually used. These are the numbers a funder or a prospective contributor would want, published whether or not they flatter the project — which, at this stage, they largely do not.

Three rules govern this page. Unavailable is reported as unknown, never as zero — a package that was never published has no download count, and showing "0" would tell you it was published and ignored. Anything derived or sampled is labelled an estimate. Every number links to the source it came from, so none of it has to be taken on trust.

9 figures are reported as unknown below.

Contribution

External contributors, last 12 months

Distinct non-bot accounts outside the org with a merged PR in the last 12 months.

External merged pull requests, last 12 months

Days since the most recent external contribution

The count and the recency belong together. One external contributor over twelve months, whose most recent work merged 173 days ago, describes a project that has received outside contribution — not one currently receiving it. The count alone would let the second reading stand.

The public repository

Contributors

Stars

Forks

Merged pull requests (sampled)

the most recent 100 closed pull requests; older ones are not paginated through

49
estimate
source

Published releases

Open issues

Days since the last public push

Distribution

Published on npm

the registry returned 404; the package is not published

not measuredsource

Published on PyPI

the registry returned 404; the package is not published

not measuredsource

npm downloads, last month

nothing is published on npm, so there is no download count

not measuredsource

PyPI downloads, last month

nothing is published on PyPI, so there is no download count

not measuredsource

Dependent repositories

GitHub exposes this only as an HTML page; scraping it would produce a number nobody can re-derive from an API

not measuredsource

Scientific usage

Public non-demo runs

requires querying the registry with demo records excluded; not attempted without a source that guarantees the exclusion

not measuredsource

Independent reproduction reports

independent reproductions are recorded as verification evidence; counting them needs registry access

not measuredsource

Both are deliberately unknown rather than counted. The registry holds demo records alongside real runs, and a usage figure that quietly included fixtures would be the worst possible failure for a project whose subject is scientific honesty. Counting them needs a query that guarantees the exclusion, and until that exists the number stays unmeasured.

How these are counted

External
Any other human account.
Internal
A GitHub account that is a public member of the ketqat organization at collection time.
Bot
An account whose login ends in [bot], or is on an explicit automation list. Automation activity is real, but it is not adoption, and counting a Dependabot pull request as a contribution is the easiest way to inflate these figures without noticing.
Known bias
Org membership can be private. A private member would be counted as external, which overstates external contribution rather than understating it -- the direction that flatters this project, and so the one worth naming.

Reproducing this page

Every figure comes from scripts/collect-oss-metrics.mjs, which reads only public GitHub and registry APIs. Run it yourself:

node scripts/collect-oss-metrics.mjs

Collected . The snapshot is committed to the site repository rather than fetched at render time, so the numbers published here are ones a person reviewed in a diff.

Want to move these numbers? Contributing lists what the project actually needs.