Repository navigation
Make MCP search results compact by default (protocol 11) - #120
Merged
Merged
Conversation
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
MCP search now returns compact hits by default, with self-contained versioned extraction references.
detail:"full"retains descriptor and diagnostic fields. Extraction accepts compact references or legacy full hits through the existing confinement and source-hash checks; query warnings count distinct confined resources/pages.Breaking: JSON protocol 11. CLI search/index/extract payload fields are unchanged except the shared version label. Installed v10 CLI extract rejects v11 documents, and strict full-result consumers must update for added fields. The docs and changelog include the compatibility matrix, binary reference specification and migration examples.
Validation: formatting and workspace/all-target clippy with
-D warningspass; workspace tests pass (635 passed, 45 ignored). All checks ran offline using the existing cached ONNX Runtime, without model loading or downloads. Independent synthetic codec vectors cover locator/metadata preservation and canonical decoding; tests cover strict selection parsing, verified extraction and protocol compatibility.The production serializer reduced 59 successful recorded payloads from 463,019 to 232,713 characters (1.990× smaller). All 63 recorded calls reconcile, including four tool errors; there are no parse/encoding exclusions. The measurement procedure and per-locator costs include a public synthetic invocation. Private transcripts/manifests are not committed. These are payload-size measurements; the paid agent harness remains deferred and no token or agent-outcome saving is claimed.