xpagesbeast
xpagesbeast

Reputation: 816

How can we store additional meta data about a Git repo?

With the growing number of microservices, and people moving around projects. It is getting difficult to find someone who supports a service.

Is it possible to put into GitHub enterprise some metadata and who the owners are and the people supporting it?

AI and other knowledge management tools such as a bot can find or index the info so it can answer.

Right now I get to the right people by word of mouth. Takes too long.

I also do not want to trigger builds.

Any ideas?

Upvotes: 3

Views: 1134

Answers (2)

BBerastegui
BBerastegui

Reputation: 313

I think that you can do that with Github custom properties!

organization administrators can create custom properties to enrich repositories with valuable information. Using these properties, you can dynamically target repository rules to apply protections on just your production repositories or to a business unit or any other way you want to classify your repositories.

https://github.blog/changelog/2024-02-14-repository-custom-properties-ga-and-ruleset-improvements/

Upvotes: 0

bk2204
bk2204

Reputation: 76649

This can be done by having some sort of standard metadata file and then having a bot crawl look in each repo and aggregate the data. This bot essentially builds a catalog of every service.

We use this approach at my current employer, where we include data about the service, its name, the team that owns it, its importance (critical, less critical, nice to have), and various other metadata. We happen to store it in a YAML file, but other options are possible. We specifically allow a file to declare multiple services, since sometimes a repository will necessarily need to host more than one logical service.

As to getting this data into the repository, we relied on people to add this data. We now require up-to-date data before deploying any service via a CI check, which means that it is less likely to be accidentally missing.

Upvotes: 1

Related Questions