Conversation
Markdown was the only ingestion path with no maximum chunk size. The post-processing loop in process_md only merged chunks that were too small, so a heading section with no nested subheading became one arbitrarily large chunk that was sent to the embedding endpoint whole, exceeding the model context window and failing the entire document upload. Cap each header section by word count before merging, apply a character backstop after merging, and clamp the embedding input as a last resort in save_chunks while still storing the full chunk text. Also make the chunk size cap unit-aware. It previously returned context_window*2 and was applied blind to word, character, and page units, which allowed a word field to be set to 16,384 words and derived a ~59,000 character image merge budget, both far past what can embed. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Add the fix write-up, release notes entries for v0.261.002, and rewrite the Chunk Sizes section and Markdown row in the admin Knowledge page so they describe the actual behavior and the per-unit caps. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Copilot Autofix powered by AI <62310815+github-advanced-security[bot]@users.noreply.github.com>
…-limit Bound Markdown chunks to the embedding token limit
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Fix governance split tab initialization
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
…admin-message Show inbound MCP disabled UI guidance
Promote Development to Staging for v0.261.002
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.
Summary
Stagingbranch tomainfor thev0.261.002release train.ENABLE_MCP_UI=trueApp Service setting is missing.Linked issue
Fixes #1362
Fixes #1364
Refs #1361, #1363, #1365, #1366
Release Notes & Latest Features
Is this visible to end users?
Is this admin-facing (Admin Settings, governance, deployment, config)?
Should this become a Latest Feature card?
Screenshot needed for the card?
Version bump
application/single_app/config.pyVERSIONthird segment bumped, or not needed because this is docs-onlydeployers/version.txtbumped, or not needed becausedeployers/was not changedTesting / validation
Staging->mainPR before creating this draft.origin/main..origin/Stagingcommit and file delta.Documentation
Security checklist
@swagger_route(security=get_auth_security()), or not applicable because this is a branch promotion PRsanitize_settings_for_user(), or not applicable because this is a branch promotion PR