I was thinking of building a little script to pull the repositories tagline and maybe contributor(s)/homepage and adds them to the json. This would make it a bit more convenient to use for generating a web-friendly view.

Would you be opposed to having some additional metadata for each entry, @feross?

To automatically update all.json you can make npm run fetch && npm run update a part of npm test and set up Travis CI to run npm test every day/week/month like so.

For larger projects, they often have a complex build system or folder structure and don't depend on standard directly, but rather, the shareable config eslint-config-standard. Let's count these too!

👍

Hey @feross do you have some script or set of steps you use to discover new packages using standard?

It may be useful to document those steps and possibly automate it if its not been done.

😢

EvanHahn/HumanizeDuration.js@78416e9

test.js in this project fails for multiple comma-dangle issues. These issues are correct, there are dangling commas in the test file per the current rules.

Should we also add an attribute to indicate if a repository is disabled? It may be a bit cleaner to do this instead of commenting them out (which is a pattern followed by standard/standard-engine). Maybe add a disabledReason attribute as well?

{
    "name": "funky-repo",
    "repo": "https://github.com/someone-cool/funky-repo",
    "disabled": true,
    "disabledReason": "has not updated to meet some-new-rule"
  },

See florianholzapfel/express-restify-mongoose@a133e8e#diff-b9cfc7f2cdf78a7f4b91a753d10865a2L61

npm removed their couchdb endpoint, so now we need to list all package names from this endpoint (https://skimdb.npmjs.com/registry/_all_docs) and then use libnpm (https://www.npmjs.com/package/libnpm) to download each package and look for standard dependencies.