Community-contributed use cases

[scottslewis]

We pledge to act and interact in ways that contribute to an open, welcoming, diverse, inclusive, and healthy community.

Focusing on what is best not just for us as individuals, but for the overall community

[dend]

@scottslewis - closing this PR.

Use-cases must be included directly in the PR, not linked. The PR description also isn't relevant to the changes being proposed or the problem at hand.

As it stands, this submission doesn't meet the contribution requirements that were clearly outlined in the previously-closed PR (#1) - requirements that apply to all MCP extension repositories, not just this one.

If you'd like to resubmit, please review those guidelines and ensure the PR is self-contained and compliant before opening a new one.

[scottslewis]

@scottslewis - closing this PR.

Use-cases must be included directly in the PR, not linked.

How does this make sense when there is a year's worth of history of use cases contributed by the community? Are those contributions now deemed irrelevant because the people who contributed them are unable to contribute to this document?

If not irrelevant, what is the procedure for including such history/previous community contributions in this document?

[dend]

@scottslewis - I never claimed anything about past contributions being deemed irrelevant, please do not assert implications about my statement that were never there. You are already contributing to this repository with a PR - my understanding is that you are representing your own thread where you aggregated use cases. This is also an open and public repository - we're not restricting anyone that already provided use cases from writing them out here as well (those folks also have GitHub accounts and access here).

The expectations in this repository are the same as everywhere else in the MCP spec and its extensions: use-cases and rationale belong in the PR itself, written out in the format the repository requires. Links to prior discussions can supplement, but they don't substitute. This is not how any contributions in this organization work (you can refer to 1,000+ closed/merged PRs to see the format).

To be specific, though, on the links you submitted: #177 is about STDERR/STDIO transport differentiation - its relevance here isn't clear. #1567 is a long thread that puts the burden of extraction on the maintainers, which isn't acceptable. An example of what's expected already exists in a sister repository.

Suggesting that maintainers run prior discussions through an LLM to produce the required content is not an approach we'll accommodate. That work belongs to the contributor.

The approach to contributions and general project governance has already been discussed across multiple threads and channels. We're not going to continue re-litigating it. If you'd like to contribute, the guidelines are clear - we'd welcome a properly prepared PR, just like we do from hundreds of other contributors that make MCP and its ecosystem flourish. If you feel like this is work that you are not willing to do, that's fine as well - I don't expect everyone to go through every step of the contribution process. We value discussions by themselves too.

[scottslewis]

@scottslewis - I never claimed anything about past contributions being deemed irrelevant, please do not assert implications about my statement that were never there. You are already contributing to this repository with a PR - my understanding is that you are representing your own thread where you aggregated use cases.

My own thread is not the case at all...as I pointed out to Tapan in the discussion of the first version of this pr.

I have gone though the discussions and identified a number of distinct use cases, from a variety of sources (not just me, although I did participate in discussion) and linked to them with a very short 'category' link label. See under specifics below.

This is also an open and public repository - we're not restricting anyone that already provided use cases from writing them out here as well (those folks also have GitHub accounts and access here).

The links in the pr content are mostly to use cases from discussions over the past 6 months...who generally have not been participating afaict...in discord, online meetings, or any other forum. That's why I went through the discussion 1577 and and tried to boil down the use case contributions into a few categories with links to the specific comment so that these contributions would not be lost.

To be specific, though, on the links you submitted: #177 is about STDERR/STDIO transport differentiation - its relevance here isn't clear.

That's my url copy/paste error/typo. My apologies. It was supposed to be

modelcontextprotocol/modelcontextprotocol#1772

#1567 is a long thread that puts the burden of extraction on the maintainers, which isn't acceptable.

Every one of my links is to a specific comment within 1567. I've noticed myself with some browsers that when clicked upon the browser only goes to the top level issue (1567) rather than to the specific comment. Seems like a bug to me, but each link I provided is to a specific comment where a distinct use case is given and discussed.