Repository navigation
Pin #1861 audit test gaps; fix Seed-OSS and LiteRT-LM Web doc wording - #1900
Merged
Merged
Conversation
Contributor
|
Chat app preview removed for |
This was referenced Nov 10, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
seed_oss_parallel_upstream.mdnow say Seed-OSS requests refuseToolDefinition.fromJsonSchematools at render, so apps validate tool name and arguments before running a call (as the 0.19.4 changelog does).ToolCall.idpassthrough and withheld content kept in the merged Web parse result (Add LiteRT-LM Web manual tool calling on @litert-lm/core 0.19.0 #1880 audit);addTearDownand an accurate name for the C API 1.0 send-failure stack test (Pin LiteRT-LM C API 1.0 send-failure text and stack (#1861 sweep 2) #1888); companion pub example selection (Test companion pub content-size check in the release verifier (#1861 sweep 2) #1889); deterministic regression test for #1892 (Fix Chrome supervisor post-done window starting before reporter_done #1893); LiteRT-LM Web limit-probe regex anchors (Report LiteRT-LM Web token-limit truncation and run automatic tool loops there #1891); MiniMax M3 exact-stream-flag guard and empty-leading-thought plus unfinished second<mm:think>and call (Take a MiniMax M3 thought only where it starts the reply, as llama.cpp does (#1887) #1894).Refs #1861 (tracker not ticked).
Production-readiness scope
lib/change.Completeness checklist
PR type guidance
Test Plan
dart format --output=none --set-exit-if-changed .test/unit/tooling(see CI)dart test -p chrome test/unit/backends/litert_lm/litert_lm_backend_web_test.dartdocker run python:3.12 python -m unittest test.ci.test_supervise_chrome_test(Linux)(cd example/chat_app && flutter test test/unit_test.dart)coverage-lib: N/A, nolib/changeMatrix Evidence
static-format-analyzeroot-vmroot-chromeexamples-testsdocs-siteMutation checks (each test fails with the mutant, passes without):
done_at = now(pre-#1893)test_post_done_window_starts_after_a_slow_reporter_pollfails 3/3 on Linux^/ without$id: nullcontent: ''exactReplyStartwithoutleadingThoughtOnlyguard (E4)takesThoughtOnlyAtReplyStart: falseforminimaxM3endsAtFirstEndTag = verbatimThoughtReview Notes
High-risk regression review