Repository navigation
Conversation
This branch has not been deployed
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Stages Agent Protocol connections behind activation gates and keeps historical B4 connections read-only. Adds Foundation-only transports, account-bound storage, migration/archive safeguards and dormant ownership controls. Relay generations retain their first accepted blocking close reason through removal.
Historical conversations retain exact drafts, opaque cached data, pending requests and unknown outcomes. Shared action guards refuse historical send/Stop/approval/local-abandon operations and dormant execution, while preserving cached reads, Check again, draft editing and current-runtime interactions. Ordered reads/events and account generations bound pending approval identity. The composer reports its actual enabled state and gives the existing read-only caption an explicit accessibility label. The five Mac smoke methods and three iOS journeys remain intact.
CI fixture repairs preserve the approved cutover and current-runtime positive controls. Owned PostgreSQL log fixtures support the persistent CI layout with provenance/privacy checks intact. Compiler controls and focused native tests run before slow aggregates. Finite diagnostic observations retain existing operations, joins, assertions, parallelism and time budgets. The optional timing observer has an explicit default-nil initializer so all five existing test-helper callers retain their construction forms. The Hub API greeting documents supported protocol 2, separately from the staged admission minimum 1.
Validation for
de350deacb5efdb049c7dd295cd36de67b95a114: source/specification, independent code-quality and final combined-source reviews passed. Fresh formatting lint passed on both changed Swift files. CI run 44219108412 is queued; compilation and runtime qualification are pending on this exact source. The latest checkpoint changes only nine initializer lines, one accessibility-label line and one API-documentation byte; all other 1,434 files are exact. Latest main7f0482ffand its published records are preserved. Repair plan and acceptance conditions.The completed E192 run is fully preserved and independently audited: 8 successful jobs, 3 failed, none cancelled. Python 44, Rust 594, runtime integration 14, both Debug builds and all three iOS journeys passed. Package and serial test compilation reported three omitted timing arguments and one inference diagnostic; native, compiler-control and model-regression execution did not occur. Mac smoke was 4 PASS/1 FAIL/0 SKIP: its earlier enabled-state, genuine editing, current-runtime and cache-preservation checks reached the later caption-label assertion at line 221. The caption had an accessibility value but no label; assertions after that failure were unreached. No missing result is credited, and these source-level repairs remain unverified until fresh CI.
The separate FB1 run retains its own results: native 41 PASS focused/default/serial, all 23 named scenarios included; selected default companions 85 PASS/1 FAIL, serial 86 PASS; dedicated physical writer 55 and publication 17 PASS; 11 compiler profiles, 33 actual controls and 11 summaries source/artifact-bound. Its serial summary was 952 tests: 949 passing methods and three pre-existing live opt-in skips. That run still failed package/Mac checks. It supplies no qualification for this newer source. Earlier failed/missing 37EDA and 402A evidence remains separate without waiver.
The original model regressions retain 9 methods/110 cases, with 8 methods/26 cases added (17 methods/136 source cases); source counts are not execution results. B17/B18 retain 23 named scenarios under the existing one-minute suite annotation, without claiming runtime budget enforcement. Native 41, selected companions 86, selected writer 53 versus physical writer 55, and broader A3/MainTen controls retain their distinct populations. C6 remains strict: 51 sources/55 inputs/52 outputs, 116 arguments, 33 vectors and 11 profiles. All actual CI, applicable focused/default/serial results, current compiler controls and final review must qualify before the existing squash merge queue; merge-group/main validation follows. PR stays draft without auto-merge or queue entry until then.
SDK publication, coordinated activation, new production deployment and release remain held. No failure-history waiver, ABI/import/freshness, physical-quiescence or live HTTP qualification is claimed.