Keep the returned value in MCP content when output was emitted - #1511
Merged
Conversation
Contributor
Cloudflare previewTorn down — the PR is closed. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A script that both
emit()s output andreturns a value only carried the return value instructuredContent, so clients that ignorestructuredContentdropped it entirely — the opposite of what theexecuteskill documented. Reported in #1509 (comment).Now the rendered return value (plus logs, as before) is appended to the tool-result
contentafter the emitted items. Skill wording updated to describe emit/return as composing rather than claiming emitted output is invisible to the model.Covers the second half of #1509; the base64/
ToolFilesandbox gap in the main report is separate.