Added links to community-contributed use cases

[scottslewis]

Motivation and Context

Provides community-contributed content.

Other questions removed for brevity since not relevant for documentation contributions (e.g. tests).

Checklist

• I have read the MCP Documentation
• My code follows the repository's style guidelines
• [NR] New and existing tests pass locally
• [NR] I have added appropriate error handling
• [NR ] I have added or updated documentation as needed

Additional context

[chughtapan]

@scottslewis Instead of linking to external sources directly, can you please describe write them down in the use-cases file and add links there. We'd want the repo to be as self complete as possible.

[scottslewis]

@scottslewis Instead of linking to external sources directly, can you please describe write them down in the use-cases file and add links there. We'd want the repo to be as self complete as possible.

The links provided already provide summaries of actual use cases. That's why I provided them...as they are actual summaries from the actual dev community, rather than someone's interpretation, abstraction or summary. Horses mouth you know.

Use cases should be grounded from actual experience...as these are. That's why I went through the discussions and picked these out...because they are from the community.

As for 'self-complete'...why?

[chughtapan]

The purview of this group currently is to gather documentation and experiences related to grouping and other ways MCP primitives should be organized. Describing all the text here and adding references allows us to iterate on things over time -- others can come in and expand upon previously discussed use-cases, describe how they have worked on it and so on. Discussions and discord channels are useful for having direct engagement but adding text to the repository requires a higher bar.

[scottslewis]

The purview of this group currently is to gather documentation and experiences related to grouping and other ways MCP primitives should be organized.

And that's exactly what these links represent: 'experiences related to grouping' from the actual people kind enough to contribute their experiences and needs...aka 'the community'.

If you want to copy/paste and/or run it through an LLM to produce text for this 'document' then go ahead but with the web I think that's silly. Especially since the history/discussion associated with the linked-to use case summaries will likely be lost, and is valuable for context.

Describing all the text here and adding references allows us to iterate on things over time -- others can come in and expand upon previously discussed use-cases, describe how they have worked on it and so on.

If what you are trying to do is have a discussion, then maybe have a public discussion...and use the history of grouping discussions and use cases over the last year as these links do...rather than starting from scratch and gatekeeping the contents...which is wasteful to the point of disrespectful of community contributions.

Discussions and discord channels are useful for having direct engagement but adding text to the repository requires a higher bar.

A 'higher bar' for community-provided summaries of use cases makes no sense to me...except as a way to exclude community contributions and previous discussions...which, I suspect, is what you are doing.

[chughtapan]

If you want to copy/paste and/or run it through an LLM to produce text for this 'document' then go ahead but with the web I think that's silly.

I'd rather not - it is important for you preserve what details you think are important to highlight.

Especially since the history/discussion associated with the linked-to use case summaries will likely be lost, and is valuable for context.

You should cite the original discussions. Those will maintain all the history.

If what you are trying to do is have a discussion, then maybe have a public discussion...and use the history of grouping discussions and use cases over the last year as these links do...rather than starting from scratch and gatekeeping the contents...which is wasteful to the point of disrespectful of community contributions.

This is a public repository. You are free to discuss. This is a new repository with a new mandate so it is expected to start from scratch.

A 'higher bar' for community-provided summaries of use cases makes no sense to me...except as a way to exclude community contributions and previous discussions...which, I suspect, is what you are doing.

I'm not gonna react to personal attacks.

[scottslewis]

If you want to copy/paste and/or run it through an LLM to produce text for this 'document' then go ahead but with the web I think that's silly.

I'd rather not - it is important for you preserve what details you think are important to highlight.

The important details I think are important to highlight are in the linked-to discussion. That's why I didn't copy it over.

Especially since the history/discussion associated with the linked-to use case summaries will likely be lost, and is valuable for context.

You should cite the original discussions. Those will maintain all the history.

That's exactly what the link does. Cite and highlight the discussion.

If what you are trying to do is have a discussion, then maybe have a public discussion...and use the history of grouping discussions and use cases over the last year as these links do...rather than starting from scratch and gatekeeping the contents...which is wasteful to the point of disrespectful of community contributions.

This is a public repository. You are free to discuss. This is a new repository with a new mandate so it is expected to start from scratch.

Right, but the document content (and history) itself is now controlled by you. If you want to create categories and put these and other use cases in (in links or sections, or whatever) then just do it.

What is this 'new mandate'? Would it be something like this? If so, please see: 'casual participation welcome'...and 'community support' and 'Facilitate discussion and knowledge-sharing among MCP contributors who share interests in a specific topic' as well as 'build consensus'. Are the previous 12 months worth of discussion, use case enumeration, experimentation and impl on grouping relevant to any of those things?

A 'higher bar' for community-provided summaries of use cases makes no sense to me...except as a way to exclude community contributions and previous discussions...which, I suspect, is what you are doing.

I'm not gonna react to personal attacks.

That's not a personal attack, it's a suspicion based upon what's happening...which appears to be that all of the community contributions...e.g. 1300 discussion, my proposal discussion, the use cases discussion is being excluded from the contents of this repo (which currently contains absolutely no content...not new, not old, absolutely nothing). If someone was to look at this repo only, they would never know about the the last year's worth of work by the dev community. Thus this pr.

And if you want things in a different format then feel free.

[chughtapan]

@scottslewis -- please see #6 for how we are planning to run this repository. Given the mandate to this group, I intend for this repository to be a self-contained wiki. I would appreciate it if you help me do that. Otherwise, I'll close this PR and incorporate the use cases myself when I get to it

[chughtapan]

Please also see https://github.com/modelcontextprotocol/experimental-ext-skills/blob/main/docs/use-cases.md as examples for where we want to get. Thanks!

[scottslewis]

@scottslewis -- please see #6 for how we are planning to run this repository. Given the mandate to this group, I intend for this repository to be a self-contained wiki. I would appreciate it if you help me do that.

Self-contained wiki? That makes little to me. Throw out all the value of linking to previous discussions, learnings, experiments, contributions?

I already have provided help to you with this pr. As I said, I went through and identified the summarized use cases described in the previous use cases and original primitive grouping proposal discussions.

All you (or claude if your model is this have to do is go to the links I've provided, copy and paste the text into claude, and claude will spit out a bullet list without any human dev community attribution, involvement, or discussion.

Otherwise, I'll close this PR and incorporate the use cases myself when I get to it

Of course you will. So much for collaboration or contributions from the dev community. But that isn't really what you are going for are you?

[chughtapan]

I already have provided help to you with this pr. As I said, I went through and identified the summarized use cases described in the previous use cases and original primitive grouping proposal discussions.

Thanks! I'll get to it.