Repository navigation
Conversation
24d9b17 to
a8de47c
Compare
[bot] 🏷 (ai: vllm-qwen/qwen3.8-27b) Conflict already resolved — no action taken this run. The merge conflict that triggered this lane was resolved by the PR author (vladkimo) before I started. He rebased the PR onto the current master ( Verified live this run:
I made no changes to the branch (nothing to fix). The PR is ready for human merge. 八问
判定:冲突已解决、CI 全绿、PR 等待人工合并(branch protection 要求 review)。CI 工程师无需进一步动作。 |
(ai: qwen3.8-27b) Reviewed the full diff against current master. The design is sound — letting the OMP host own truncation recovery (instead of the proxy burning a second retry budget) and keeping transport diagnostics local are both the right calls. The marker-header plumbing, the egress strip, and the diag-field additions are all clean. One regression found and fixed (pushed to your branch): the Also: your rebase onto master already resolved the only merge conflict (CHANGELOG), so the PR is mergeable. All targeted suites pass (178 tests / 170 pass / 0 fail / 8 skipped); typecheck clean. 结构化八问
收工判定:commandcode 回归已修并推送到 PR 分支( |
[bot] 进度汇报:工作已全部完成,进程卡住发生在收尾之后,无未竟事项。 已完成:
卡住原因:工作完成、review 已发布后,进程在收尾阶段挂起(疑似网络/工具调用挂起),非阻塞性。重启后状态完整,无需补做。 收工判定:PR 现可合并(待常规 review/check 门禁)。 |
[ework] CI fix verified green ✓ |
1 similar comment
[ework] CI fix verified green ✓ |
|
Landed via batch PR #6482 (merged) — closing as superseded. |
910bc13 to
2153152
Compare
Summary
Follow-up to #5956 (merged 2026-11-09, closes #5951). After that merge the proxy still spends its own one-shot replay when the requesting host already owns truncation recovery, so one cut can be recovered twice — once by the proxy and once by the host's own retry.
src/util.ts,src/agent/omp-native.ts: the OMP native bootstrap stampsx-bili-host-recovery-owner: ompon the requests it sends through the proxy.src/server/headers.ts: the marker is stripped before upstream egress, next to the other internal hop headers.src/plugin.ts,src/server/relay.ts: both pipe entry points read the marker past the newstreamErrorShapeparameter and passhostOwnsTruncationRecovery; when it is set, a reasoning-only EOF / read error / zero-byte cut is reported to the client in-band without spending the proxy retry. Pending answer filters, visible answer text, tool calls and an exhausted budget still refuse the replay.src/stream-error.ts:TruncationDiaggainsreadErrorCause,upstreamRequestId,upstreamLogId,upstreamCfRayandreplay; themetafield is removed from the client-visible terminal SSE frames, so transport causes and upstream trace ids ride the proxy log line only. Codes and messages are unchanged, and thecommandcodebranch from [dsh] Retry-exhausted degenerate turn is serialized as successful completion (0.1.213) #5958 keeps its shape.Wire compatibility
The only client-visible change is the removal of one diagnostic field from terminal error frames:
data: {"error":{...}}+data: [DONE]error.metaevent: error{"type":"error",...}meta{"error":{"code":503,"status":"UNAVAILABLE",...}}error.metaevent: error{"type":"error","error":{...}}error.metaThe google frame keeps
code: 503andstatus: UNAVAILABLE; it does not carryupstream_stream_truncated, so nothing it said is lost.Evidence
The recovery-ownership matrix, measured on the same reasoning-only cut in
tests/issue5951-reasoning-only-cut-replay.test.ts:x-bili-host-recovery-owner: omptests 170 / pass 162 / fail 0 / skipped 8across the eight files this PR touches (issue5951-reasoning-only-cut-replay,issue2171-zero-byte-truncation-retry,issue2328-terminal-diag,stream-error,proxy-stream-error,omp-native,claude-native,native-intercept-selfheal).npx tsc --noEmit -p .clean;npm --prefix kernel run buildclean.c26c31ac),npm test:tests 11017 / pass 10951 / fail 4 / skipped 62 / todo 0in 557 s. The four failures are host artifacts, not this change:findProjectJsonlFilesandwriteDshClientShimFiles (#1590)compare/tmpagainst/private/tmp, plus the two#5627cases — the same four reproduced on the pre-rebase base3ac3e95cin a clean-env four-file control (pass 389 / fail 4on both trees, identical names). No full-suite control was run againstc26c31acitself. A#3544stale-proxy-rotation case failed once in a pre-rebase run on this machine, does not reproduce at this HEAD and passes in isolation on both trees — unattributed.bun run tests/e2e-omp-socket-cut-smoke.mjs(harness not included in this PR) against the oh-my-pi source checkout — one cut produced2upstream requests,hostCalls 2, oneauto_retry_start/auto_retry_end success, final assistant textRecovered response., and a proxy diag withreplay.replayAvailable=true,openStreams=0. That cut was plain text, not reasoning-only, and it ran on the source checkout (v18.18.2), not the installed 18.18.8 binary.How the reasoning-only predicate is covered
Two cases in
tests/issue5951-reasoning-only-cut-replay.test.ts: "the host owning truncation recovery suppresses the replay" (read error after reasoning-only output →refetches === 0, sentinel still emitted, the partial reasoning still forwarded once, diag recordsreplayAvailable: true) and "the host owning recovery also suppresses the zero-byte replay". The sibling case on identical input without the marker assertsrefetches === 1, so the suppression is not a tautology. The marker itself is covered at the stamp (native-intercept-selfheal), the egress strip (claude-native) and the client-header source (omp-native).No historical baseline run is claimed: the cut cannot be produced against an installed host from a test, so the control is the pre-marker behaviour asserted by the sibling case in the same file.
Docs: no page describes the removed
metafield —docs/architecture/pipeline.md:30already describesstream-erroras a "minimal SSE error + finish", this change makes the frame strictly more minimal, and:28already states that internal hop headers are stripped at egress, which the marker joins.Merge Danger
Door: two-way
Blast Radius: wire
Closes #5951