SEP Guidelines: Require reference implementation prior to formal review #2086
+13
−10
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Update
sep-guidelines.mdxto require reference implementation prior to formal review.Motivation and Context
This PR arises from a discussion in the maintainers' channel of the MCP Contributors' Discord.
From the current SEP Guidelines:
It would be beneficial to require reference implementations prior to SEP acceptance. Very simple ones might be safe to evaluate without it, but SEPs that have a lot of impact to the spec (e.g., Tasks, MCP Apps) should have a reference built and worked with / evaluated by the maintainer community prior to acceptance.
In the case of MCP Apps, Goose and MCPJam have independently built support for the proposed spec prior to adoption. This will provide a a feedback mechanism so that perhaps when it is adopted, it will be in its best shape.
Of course, all SEPs will not attract unprompted community implementations prior to the spec's acceptance. By requiring the reference first, at least the proposer will have a chance to work through their ideas and optimize the SEP for real world use. And the core maintainers can see what the actual implementation implications will be prior voting on it.
For instance, the Server Variants SEP seems like a good idea, and promises a reference implementation upon acceptance. But it is sufficiently complex that some actual testing should be done to see if it will actually work, and if it is optimal.
Otherwise, we end up with spec changes based upon a proposal alone and when real world issues show up in implementation, there's no easy / clean way to undo it or modify it toward its optimal shape.
How Has This Been Tested?
Served docs locally, and proofread.
Breaking Changes
Nope.
Types of changes
Checklist
Additional context