Skip to content

[PR] fix: merge current master into ref-floor parity branch - #6197

Closed
Maziar123 wants to merge 742 commits into
ranxianglei:2026-11-03_ref-floor-digest-parityfrom
Maziar123:2026-11-03_ref-floor-digest-parity
Closed

Maziar123 wants to merge 742 commits into
ranxianglei:2026-11-03_ref-floor-digest-parityfrom
Maziar123:2026-11-03_ref-floor-digest-parity

Conversation

@Maziar123

Copy link
Copy Markdown
Contributor

Purpose

This maintenance PR merges current master into the head branch of #5597. It updates the existing contribution; it does not replace it.

Conflict resolution

The only conflict was in src/state-hydration.ts. The resolution preserves both independent fixes:

The resulting diff against current master contains the same nine source and test files as the original PR.

Verification

  • npm run build — passed.
  • npm run typecheck — passed.
  • Focused ref-floor, orphan-journal, hydration-field, and unified-storage tests — 25 passed, 0 failed, 1 optional better-sqlite3 case skipped because that native dependency is not installed.

The full test suite was not run locally. Remote CI on #5597 remains the authoritative full-suite check after its head branch is updated.

Maintainer action

Merge this maintenance PR into 2026-11-03_ref-floor-digest-parity. This will update #5597 with current master and its resolved conflict.

ci-check and others added 30 commits November 9, 2026 00:36
…surface other bucket (ranxianglei#5963)

Review fix on the context-analysis feature: the idle-gate 'distill' bucket
matched "no tier mass-qualified" / "summary-pool|distillation" — strings the
kernel never emits in an idle reason (checked against the reason construction
in kernel/src/compress.ts). The counter was permanently 0 on real traffic while
count-ready T2 idles silently fell into the unshown catch-all bucket.

- drop the two dead patterns and the always-zero distill bucket (T2/T3
  readiness is already reported from lastReadiness in the verdict);
- show the other count in the report line so unclassified idles are visible;
- test rows now use the kernel's actual reason forms (incl. the full
  ', blocked: T1 (cadence)' suffix and a count-ready T2 idle).
…canonical port (ranxianglei#5955)

The attach/rotation decision table had one uncovered cell: a same-lane
resident on the canonical port running DIFFERENT bytes whose health probe
never answers (BUSY: TCP connect succeeds while the event loop stays inside
long sync slices, e.g. a multi-minute fold on an oversized session). Attach
vetoes foreign bytes (ranxianglei#1225), ranxianglei#4125's upward-adoption veto-lift needs an
ANSWERED version a BUSY holder cannot give, ranxianglei#3544 rotation excludes
version-unverifiable residents by design (ranxianglei#4811), and the spawned child waits
5s behind a same-lane predecessor then ladders away (+1 sticky drift) or dies.
No actor ever resolved the holder, so the port stayed wedged until the process
died on its own — which the event-loop watchdog only guarantees for >=5min
CONTINUOUS blocks; the saturation class never trips it. Oversized sessions
then 422'd indefinitely on the stuck lane.

Resolution: bounded patience across bring-up cycles, then a supervised forced
takeover. A persisted observation ledger (state dir, atomic write, keyed
origin#pid#birth-stamp so a REUSED pid starts its own window instead of
inheriting the dead occupant's history) accumulates sightings across client
restarts. After WEDGED_TAKEOVER_WINDOW_MS (10min = 2x the watchdog hard
timeout) of continuous observations across >=3 distinct cycles, the launcher
requests a graceful shutdown, terminates if the listener survives, hard-kills
as last resort, verifies the listener closed, gives kernel socket state a
settle after an abrupt kill, and continues to the normal spawn. Any doubt
falls through to the untouched spawn ladder — never worse than today.

Adjacent surfaces deliberately unchanged: ranxianglei#3544 rotation (healthy older
holders), ranxianglei#4125 upward adoption (healthy newer holders), ranxianglei#4811 same-lane
attach (matching bytes), user-zone and foreign-lane holders all keep their
existing paths. killAbandonedChild keeps its exact behavior (group kill +
handle-level SIGKILL) via the new shared hard-kill helper.
The PR bumps kernel/package.json 0.0.188 -> 0.0.189 (required by the
Kernel Boundary for the kernel/src/wire/google.ts change) but left both
lockfiles at 0.0.188. Regenerated with npm install --package-lock-only
(matching the lock-sync CI job) so the embedded kernel entry in the root
lock and the kernel lock's own version line agree. Mechanical lockfile
sync only — no dependency or source change.

Fixes ranxianglei#5971
…leanup)

Line 111 asserted st.overflowArmCompletionReserve === 5000, identical to the
pre-existing line 136 (ranxianglei#3450). Removing the redundant copy; coverage is
unchanged.
… turn (ranxianglei#5604)

A preflight fold of the tail (soft protection relaxed to 0, or a tail tool
pair in neverPreserveRecentTools) can leave the rebuilt Anthropic messages
ending on the surviving assistant turn. Anthropic and Bedrock reject that as
assistant prefill ("This model does not support assistant message prefill",
"final assistant content cannot end with trailing whitespace").

prepareAnthropic now closes such a tail with one thin ephemeral user nudge,
inserted before the trailing system run and after all cache-stable content,
via the kernel's insertBeforeTrailingSystemRun. A client-authored assistant
tail (deliberate prefill) passes through unchanged.
…ignature, not a string no receipt carries

The re-pin asserted !second.includes("Eligible ranges") — that capitalized
phrase appears in no receipt (the recovery hint reads "Current compressible
ranges", the one-liner lowercase "eligible ranges"), so the check passed
vacuously. Pin the actual lecture signature instead: "compress takes ONE
'content' STRING" is present in the full receipt and absent from the
one-liner, so a regression that re-expands repeats into the full lecture is
caught.
wedgedLedgerFilePath() is referenced only inside launcher.ts — the
deps-injection fallback at line 5691 — while tests supply their own path
via deps.wedgedLedgerPath. The leading `export` made it INTERNAL_ONLY under
scripts/check-unused-exports.mjs. Drop it to match the sibling module-local
ledger helpers (isWedgedLedgerEntry, readWedgedLedger). No behavior change.
…ile name

ranxianglei#5392 is an unrelated open DSH pruner bug and ranxianglei#5585 an unrelated compaction-leak
issue; the real source threads are ranxianglei#5390 (Claude-through-Google tool_result-first
part order) and ranxianglei#5584 (codex marketplace .tmp containment). Comments and the
test file now cite the verified GitHub numbers.
…trine ranxianglei#496/ranxianglei#3909)

The band's premise "any parseable body over 32 MiB is >=~8M tokens" holds
only for TEXT-dominated bodies: base64 pixels are excluded from the token
meter by design (ranxianglei#496/ranxianglei#1800/ranxianglei#857), so an over-band IMAGE body is not
automatically over-window. ranxianglei#3909's ~48 MiB image-dense chat requests would
have been refused 413 stage=ingest on every turn by the default-on band —
breaking a supported workload the outbound shed exists precisely for.

- handle.ts: over-band refusal now requires countImagesInRawBody(...) === 0
  (probe-gated, so text-dump rejection stays parse-free). The Responses-lane
  exemption is unchanged.
- tests/issue5794-ingest-cap.test.ts: +1 e2e — an over-band body carrying
  two ~800 KB images must NOT be refused at ingest (200 + forwarded), on a
  1M window where the pixels stay token-fit.
- yaml/lifecycle/CHANGELOG: wording corrected to "text-dominated"; docs
  regenerated (gen-config-docs).
… alarm (ranxianglei#5955)

Acceptance run on Windows (issue reporter, branch @ 33bec44) passed all
three points but measured ~12.4s of pure budget spin in the escalation
ladder against a structurally unresponsive holder (the port released in
~250ms after the hard kill), plus a false-alarm warn on every successful
win32 takeover:

- Escalation stages tightened from 3s+5s+2s+5s to 1s+2s+1s+2s. The
  shutdown-request timeout is now parameterized (requestInstanceShutdown
  gains an optional budget; ranxianglei#3544 rotation keeps its 3s default).
  Worst-case release verdict drops from ~15s to ~6s (+2s post-kill settle
  unchanged). The graceful-first discipline is kept: after a 10-minute
  health silence it is insurance, bounded to at most ~3s of cost.
- hardKillProxyTree win32 fallback no longer warns when the direct SIGKILL
  throws ESRCH — taskkill already reaped the tree, which IS the goal;
  mirrors the POSIX branch suppression.
…-carry regressions

Ported from parallel PR ranxianglei#5979 (same fix for ranxianglei#5968; being consolidated into this PR): three unit cases not covered by this branch - applyForkAdoption deep-copy isolation (mutating the adopted child never leaks into the source session), the straddled counter for partial spans while the block is transiently inactive in the source, and covered-ref carry into the child numbering.
…ladder

The field-tuned timings in 2826c51 shipped without tests: the escalation
budget (1s POST + 2s post-shutdown wait + 1s SIGTERM drain + 2s post-kill
wait + 2s settle) and the win32 ESRCH silence are unguarded, so the next
re-tune can silently bring back a ~12s stall or emit a false "listener may
survive the cleanup" alarm on every takeover.

tests/issue5955-wedged-same-lane-takeover.test.ts (18 cases):
- candidate matrix: same lane + canonical port + BUSY + fingerprint
  mismatch is the ONLY cell that may be taken over
- ledger boundaries: window, min observations, 300s stale gap, persistence
  across client restarts, fresh window after pid reuse (birth stamp key)
- escalation order shutdown -> SIGTERM -> hard kill, and the short BUSY
  budget instead of the 3s graceful default (staging bounded to ~6s)
- kill failure falls through to the normal spawn ladder, entry kept
- port:0 lazy sticky-lane resolution; graceful path never signals/kills
- negatives: healthy older/newer/same-build BUSY holders, user zone,
  foreign lane, other port and explicit attach targets are left alone
- win32: a pid already reaped by taskkill stays silent, real failures
  still surface (ranxianglei#4125 vs ranxianglei#5955 field noise)

Exports hardKillProxyTree so the win32 silence assertion can call it.
…nsit (ranxianglei#5981)

An empty compress call has two completely different causes that are
bytewise identical once the parser sees them: the model wrote an empty
companion call (ranxianglei#1366), or the argument deltas never arrived because the
upstream stream was cut before its terminal event. The receipt assumed
the first and told the model not to re-issue, so a transport failure
became an abandoned compression.

Log evidence (bili.log, two incidents, one session each): the truncation
warn lands, then 37-115ms later compress is dispatched with
`kind=missing-content invalidItems=0 keys=[]` — zero argument bytes ever
reached the proxy. No parser salvage can recover bytes that never
arrived, so the fix belongs at the decision point that advises the model.

Stamp the fact where it is observed (the four emitUpstreamTruncation
sites) and consult it in applyRanges before the ranxianglei#847/ranxianglei#4812 ledgers, so a
transport failure is neither charged to the model's repeat-failure
breaker nor collapsed into a do-not-retry refusal. The stamp is
non-consuming with a 5-minute TTL: one cut stream may dispatch several
calls, and a stale stamp must never re-label a genuinely empty call.

Without a stamp the ranxianglei#1366 receipt is unchanged.

Fixes ranxianglei#5981
…er (ranxianglei#5981)

The first cut of this fix skipped BOTH failure ledgers, which left the
re-issue advice unbounded: a transport that stays broken would let the
model retry a dead stream until the turn dies — exactly the runaway the
ranxianglei#2146 session streak exists to stop.

Split the two ledgers by what each one actually means:

- ranxianglei#847 (compressFailKeys) stays clean. It tells the model "you have now
  requested this exact spec N times", which is false here — the model
  never sent a spec; the bytes were cut.
- ranxianglei#4812 (compressFailStreak) IS charged, on a dedicated
  lost-in-transit spec label. A persistently broken transport must still
  bound retries, and at the threshold this collapses into the same armed
  refusal every other receipt gets.

Adds the negative direction: a single lost-in-transit refusal still
carries the re-issue advice.
…st-in-transit branch

The lost-in-transit branch gates on isEmptyCall; the first pass pinned only
the empty side. This row pins the other direction: a shape-drifted NON-empty
invalid call ({ranges: …}, ranxianglei#362) with a live truncation stamp must keep its
own receipt (no valid ranges parsed) and still charge the ranxianglei#847 spec guard -
bytes present means the transport story does not apply.

Additive test-only review commit on PR ranxianglei#5982 (issue ranxianglei#5981); no source change.
…ianglei#732 follow-up)

A thinking model can end a turn with reasoning + whitespace-only text
(e.g. "\n\n") and finish_reason=stop. No tool call, no visible answer,
client stalls. The ranxianglei#732 one-shot retry gate read this as 'visible output
already reached' (length-2 text) and skipped the retry, passing the
empty turn through silently.

- track non-whitespace visible text (forwardedVisibleText) separately;
  the ranxianglei#732 gate now uses trimmed text + that flag, so whitespace-only
  terminals take the retry path exactly like pure-reasoning terminals
- the ranxianglei#1988 truncation gates keep the original forwardedVisible
  semantics (forwarded, not parsed) — untouched
- observability: warn when the one-shot budget was consumed and the
  retry output was empty again (double-degenerate), so the fleet can
  count the residual case the host nudge still covers
- compress receipt hardening: the FAILED receipt now ends with an
  explicit return-to-task instruction, reducing the observed pattern of
  a model derailing into the compression bookkeeping after a refused
  call and then going quiet
…anxianglei#5963)

- pairing used pendingT1 or fell back to the whole-session token count, so a T2/T3 offer read as under-harvested even when the model fully executed it; use the injected tier's own mass (pendingT2/pendingT3) with the same fallback

- regression test: a T2 offer of 80K fully folded reads ratio 1 (old code: 0.2)
…st-loop

[PR] Compression cycle fix: identify and confirm pointer echo, condense duplicate failure receipts into a single line, and co | 压缩循环修复:认出并确认指针回声、把重复失败回执缩成一行、修复折叠后压力表虚高
…fold-summary-v2

[PR] Put each compression summary at its fold point; the old slot keeps a pointer instead of a full compress command (compressAtFoldPoint v2)
…te-retry gate (ranxianglei#5986 review)

Review finding 1 of PR ranxianglei#5986: the flag was provably equivalent to the gate's
assistantText.trim().length === 0 check — both reset per attempt at the top of
the retry loop, and assistantText is mutated only by delta accumulation, so the
flag could never disagree with the trim check. Removing it keeps one source of
truth; behavior is unchanged. The truncation gates keep forwardedVisible, which
needs forwarded-not-parsed semantics; that boundary is now documented on the
trim condition itself.
…scoped budget (ranxianglei#5986 review)

Review finding 2 of PR ranxianglei#5986: degenerateRetried is declared once per request
(core.ts top of runCompressLoop) and shared with the ranxianglei#2303 draft-tail retry
path — it is never reset per round. The warn's comment and log line both
claimed 'consumed this round' / 'the retry output was empty again', which
misattributes the pass-through in the chain where an earlier round (or a
draft-tail retry) spent the budget and a LATER round degenerates fresh. Reword
comment and log message to the accurate request-scoped attribution; keep the
stable 'DOUBLE-degenerate' token for fleet grep continuity. No logic change.
…ei#5986 review)

Review finding 3 of PR ranxianglei#5986: the double-degenerate pass-through (per-request
retry budget already spent, round still ended degenerate) was observable only
in fleet logs — no test captured loggerLog. Add R1/R2 via the setLogCapture
seam: R1 drives the degenerate-re-request + degenerate-retry chain and asserts
exactly one DOUBLE-degenerate warn with the request-scoped attribution; R2
drives a successful retry and asserts NO double-degenerate warn while proving
the capture is live via the auto-retry warn. Gate discipline holds both ways.
…re receipts (ranxianglei#5986 review)

Review finding 4 of PR ranxianglei#5986: the 'return to the original task' sentence was
appended only to the 'no valid ranges parsed' receipt, though the observed
failure mode (model entangled in compression bookkeeping after a rejection,
then stopping out) applies to every parse-failure receipt. Extract it into a
named constant RETURN_TO_TASK_SUFFIX and append it to the empty-call and
argument-corruption receipts as well; the armed-breaker one-liner already
carries its own 'Continue your task without compressing.' and is left
untouched. The policy-refusal receipts (superseded-ref, out-of-recommendation)
are deliberately unchanged: their repair instructions ('re-issue with updated
labels', 'stay within the recommendation set') take precedence over 'move on'.
…-lane-small-fixes-batch

[PR] fix(codex): Codex lane minor fixes roundup ranxianglei#3 — WS error object + marketplace containment + Google part ordering | fix(codex): codex 车道小修合集第三辑 —— WS 错误对象 + marketplace 收容 + google part 序
…xt-analysis

[PR] feat: Context Analysis — init size / injection frequency / compression acceptance rate / diagnostic suggestions (ranxianglei#5963) | feat: 上下文分析 — 初始化多大 / 注入多频 / 压缩接受率 / 诊断意见(ranxianglei#5963)
…count-ingest

[PR] Hard-block session history over-limit along real dimensions: message-count gate + 32MiB ingestion cost band + fatal erro | 会话历史超限按真实维度拦死:条数闸 + 32MiB 摄入成本带 + 判死信息点名超重字段 (ranxianglei#5000 ranxianglei#5794 ranxianglei#5885)
…truncated-upstream-compress-receipt

[PR] fix: tell the model to re-issue a compress whose args were cut in transit (ranxianglei#5981)
ranxianglei and others added 25 commits November 10, 2026 21:22
…r-user-slot-fail-open

[PR] fix: fail open when a marker strip would blank the latest user turn (ranxianglei#5812)
…aders-merge-fix

[PR] fix(agent): pass merged headers to pfRequestHeaders so watermark count is emitted (ranxianglei#5813)
…-summary-floor

[PR] fix: judge summary floor at batch level for multi-range compressions; expose minSummaryLength in panel
…-field-diag

[PR] fix(storage): add field names to drift diagnostics + disable incremental fast path during drift latch (ranxianglei#5677) | fix(storage): drift 诊断补字段名 + 漂移锁存期间停用增量快速路径 (ranxianglei#5677)
…e-timestamp-ledger-carry

[PR] fix(resume): inherit timestamp ledger so resumed sessions re-render provider-cached tag bytes
…ndle-ca-only-filter

[PR] fix: keep only CA certificates in combined-ca.pem (ranxianglei#5802)
…route-env-on-unload

[PR] fix(hermes-plugin): restore routed process env on unload and watchdog exit (ranxianglei#4644 follow-up)
kernel version: 0.0.183 (PR) vs 0.0.192 (master) -> 0.0.193
…anxianglei#6132)

Owner-directed full removal of the experimental compressAtFoldPoint /
compressAtFoldPointClearThinking feature (v1 ranxianglei#5471 + unreleased v2 ranxianglei#5949):
two distinct live pathologies in 9 days on its own target model family —
2026-11-08 voluntary re-compress loops (v1, ranxianglei#5949 field report) and
2026-11-10 authorship misattribution (v1 in v0.1.213 production: the pair's
relocation left the turn's own answer text bare at the transcript tail, and
a 27B model read its own last utterance as a user message, m00376 chain).

- delete kernel/src/relocate-compress.ts + kernel/tests/relocate-compress.test.ts
- drop the relocate-compress-carriers pipeline node (after hide-compress-calls,
  before view-latch) and its registration
- drop Config.compressAtFoldPoint / .compressAtFoldPointClearThinking /
  .reasoningDroppable and the public relocateCompressCarriers/plan exports
- drop the already-folded-echo parse kind + v2 pointer marker
  (parse-compress-input.ts) and the two echo tests
- kernel 0.0.192 -> 0.0.193 (both package-locks)

kernel-change: removed relocate-compress.ts, the pipeline node, 3 Config
fields and the public exports (full feature removal, owner decision after the
2026-11-10 incident, ranxianglei#6132); with the switch off the transform was a
byte-identical no-op, so removal changes nothing for the default layout.
kernel 0.0.192 -> 0.0.193.

Closes ranxianglei#6132 (kernel side; host/web/docs follow in the next commit of this PR)
…#6132)

Companion to the kernel removal: strip the experimental fold-point
relocation feature (v1 ranxianglei#5471 + unreleased v2 ranxianglei#5949) from everything else.

- env knobs BILI_COMPRESS_AT_FOLD_POINT(_CLEAR_THINKING) + resolvers
  (knobs.ts), file keys compress.compressAtFoldPoint(_ClearThinking)
  (config.ts: CompressSettings fields, COMPRESS_SETTING_FIELDS, parser
  loop), compress-settings.ts cascade/pass-through/reachability,
  relay.ts round-2 mirror + drop-predicate install, the four prepare*
  install sites, plugin-render.ts S4 mirror, makeFoldPointReasoningDroppable
  (server.ts), the v2 already-folded-echo ack in stream.ts
- stale configs: one-shot INERT deprecation warning extended to the two
  removed keys (every level incl. top-level), values ignored
- web config row + i18n strings (en/zh), config-reference yaml, site docs,
  site i18n strings, blog post + article entries deleted (owner instruction)
- docs configuration en/zh + lifecycle ledger (rows removed, §5 note),
  CHANGELOG entry
- tests: delete tests/compress-at-fold-point.test.ts, drop the echo test
  from compress-reject-reasons.test.ts, extend removed-compress-keys-inert
  with the removed fold-point keys

Closes ranxianglei#6132
…5765 tool surface

The acp_retire/acp_restore/acp_rewrite schemas are always-on in every
plugin manifest, so the pinned manifest/fork/launcher tool lists, the
insertion-order lists, and the codex-official prompt regex ('four' ->
'seven' context-management tools) all needed the three new names.

Fixes the 9 stale-assertion failures across plugin-protocol, plugin-fork,
launcher-plugin-mode and codex-official (deterministic on PR head, green
on master).
… surface

The four committed storage-render snapshots are byte digests of the
forwarded wire bodies; the always-on retire/restore/rewrite tool
schemas change every body (+2685B on the chat wire per turn), so the
digest/bytes fields were refreshed via
scripts/update-storage-render-goldens.ts. Structural fields (turns,
summary placement) are unchanged. Fixes golden-storage-render +
storage-engine-proofs (8 tests, 4 wires).
…y growth

The always-on retire/restore/rewrite tool schemas add ~0.66-0.7k
est-tokens (~2.6KB) to every forwarded body. Several e2e scenarios are
deliberately calibrated a few hundred tokens under their trigger
boundaries (the repo's established pattern when the wire surface
shifts: ranxianglei#2750 wording change, ranxianglei#3298 fold-boundary retune, ranxianglei#1933
bill-rate retune), and the constant body growth pushed them over:

- e2e-relay-5xx-selfheal: tool_result 12000 -> 11700 chars keeps the
  payload at the intended ~97% of the 32k window.
- issue1933 F2: 1.4x/15x4000 -> 1.5x/14x4000 keeps the billed total at
  exactly 21000 (the scenario's +1k-above-target) while the estimate
  lands ~1.5k under the 20k trigger.
- ranxianglei#663: repeat(235) -> repeat(220) keeps the post-fold forward at 4
  messages (the mock only misclassifies exactly-1-message bodies).
- e2e-google-compress: clamp the mock's fold endId to refIds[max(len-6,4)]
  — with the +2.6KB body the 24KB demand gate now opens at 12 refs, and
  the len-10 tail then degenerates to a single-message range
  (463 chars < minCompressRange, hard refusal). The len-6 clamp also
  fixes a steady-state ratchet each fold out-earned by ~500B, so the
  upstream bodies stay bounded (client 65.7KB vs peak 31.5KB, ratio
  2.09, assertion requires >2).
- ranxianglei#280: repeat(235) -> repeat(155) — the overflow-relaxed preflight fold
  now eats 3 more messages before the hard protection floor; 155 (the
  measured pass window is 100..215) keeps MARKER_8..11 in the forward.
- ranxianglei#1843: mock bill 8000/6617 -> 8657/7274 (a real encoder bills the
  +657-tok schema text too) so the learned per-image cost lands at
  1813.5, inside the (1500, 5000) sanity bound.
- preflight-hold ranxianglei#568: repeat(235) -> repeat(220) (same family as ranxianglei#663).
- issue2672: idx8 2562 -> 207 chars (it is BLOCKED — no compressible
  slot — so its size never touches the walk geometry) and tail 358 ->
  294 chars (74 tok each keeps the 1000-tok recent-zone walk stopped at
  idx9 exactly as on master: 5x74+5+675 = 1050 >= 1000). Restores the
  pre-ranxianglei#5765 post-fold payload (~7.6k tok vs the 8k window).
- cache-proof host-compaction: seed FILLER 6 -> 5.5KB/turn — the
  +2.6KB schemas pushed the second-phase seed body one body earlier past
  the 64KB demand threshold (66.9KB vs 65.5KB at body ranxianglei#11), landing the
  seed fold one turn ahead of the ranxianglei#2695 geometry the zombie-settle
  bound (9) and the rearm-shrink pin were measured against; 5.5KB
  restores the master-aligned fold turns (verified via PROOF_DUMP:
  seed folds at 8->9 and 12->13 on both, rearm at 21->22).

All 15 affected files green (156/156), matching the 27 deterministic
CI failures 1:1; the identical suite is 155/0 on master.
…nxianglei#5765)

The ranxianglei#5924 architecture-docs guard (landed on master via the branch sync)
requires every tracked src/ file in reference/architecture.md; the new block-ops
host executor module was added by this PR without its map row, so the guard
failed on the merged head. One-line registration next to the acp-* cluster;
guard passes locally (320 src files covered, no ghosts).
…+ tag-stripped body (ranxianglei#6160)

The scripted queue key was conv|sourceText.slice(0,64) over the RAW last-user
text, whose ACP render tag is volatile by design: the post-fold ledger
alignment (ranxianglei#5608-D defers restampRefsToLedger to the next successful fold)
renumbers old refs (m00001 -> m00015, refs-never-reused) and raises the
ranxianglei#5608-C time-attribute high-water mark, so the very next request re-renders
the same message with time= dropped. The 64-char key slice contains the tag,
so one scripted turn split into two queues the instant a fold landed and the
directive queue replayed from scratch — the ranxianglei#2791 tests' deterministic
"compress must be CALLED exactly once (got 2)" on local node-25 and the
intermittent red on CI (same SHA red and green across runs; the alignment
no-ops when the unified store has no visible rows yet at the post-fold
scheduleSave).

The key is now conv|<source ref>|<stripAcps(source).slice(0,64)>: the ref
names the message occurrence (stable for the turn; a remint voids the script
anyway) and keeps byte-identical prompts distinct (switch/back both send
请调用probe_tool); stripAcps (existing helper) drops the volatile wrapper.
One site (answerFor) covers all six suites sharing the fake.

Verified locally (pi-stable 0.83.6): both ranxianglei#2791 tests green — the fold run's
queue continues through the fold (probe -> compress -> acp_status, one call);
full pass over native-pi/dsh/opencode/delegate-notify/subagent-sessions
(the two remaining local failures reproduce identically on the unmodified
fake: pre-existing local-environment artifacts).

Closes ranxianglei#6160
…e-compress-at-fold-point

[PR] remove: the experimental compressAtFoldPoint feature — fold-point relocation (v1 ranxianglei#5471 + unreleased v2 ranxianglei#5949), ranxianglei#6132
…ei#5984)

dsh-live-compact is a competing compaction engine that takes over dsh's
stock compaction row (its bundled cordis.patch.yml disables
compaction-basic and registers its own engine row; its preset takeover
renames the preset's compaction row to itself), so bili's load-time
compaction-disable overlay (ranxianglei#3354), which recognizes only the stock row
names, is structurally blind to it. Its checkpoints can land outside
bili's knowledge and destroy the compression substrate - the ranxianglei#5984
session.

- scanDsh: dsh-live-compact entries (bare name / npm spec / bounded path
  segment, isDshLiveCompactEntry mirroring isOpencodeAcpEntry's
  discipline) are now a deterministic "known" finding (knownId
  "dsh-live-compact") instead of a keyword [suspected] one, so the
  conflict ledger / acp_status / web banner surface it as a confirmed
  conflict; lookalikes stay at most keyword-suspected
- announcements/remote.json: web notice (ranxianglei#5548) for dsh profiles with
  the plugin, with the removal recipe (dsh plugin remove
  dsh-live-compact) and a CTA to ranxianglei#5984
- docs: known-conflicts list updated (configuration.md / zh-CN)
- tests: 3 new cases + detail-shape / not-benign / not-sibling /
  not-display-only assertions
….blockOps (default off)

Owner ruling (review of this PR): acp_retire / acp_restore / acp_rewrite
must sit behind ONE switch, default OFF, injected only when ON — the
always-on surface costs ~0.7k estimated tokens per request on every
session of an enabled proxy.

Follows the absorb opt-in pattern end to end:
- kernel: definitions + name consts stay exported, but the 3 tools leave
  the default arrays (ACP_TOOLS_{OPENAI,ANTHROPIC,RESPONSES,GOOGLE},
  ACP_READONLY_TOOLS_RESPONSES, ACP_TOOL_NAMES); they stay in
  ACP_MUTATING_TOOLS (dispatch must drive the compress loop when enabled).
  New BLOCK_OP_TOOL_NAMES set + buildBlockOpsSystemPrompt(lane) section
  builder; the default prompts are back to "four context-management
  tools". New Config.blockOps { enabled } (off by default); absorb/ccr
  tool-name predicates include the block ops.
- host: compress.blockOps config (3-level merge like absorb);
  blockOpsEnabled / effectiveBlockOpsConfig / storeEffectiveBlockOps
  (per-session, plugin tool API parity); isProxyToolFor, the proxy-lane
  inject* extras, the responses hybrid lane, the round-2 cache seam, the
  plugin manifest + tool API, the MCP dialect note and the system prompt
  all inject/advertise the 3 only while enabled; the MCP note, prompt
  and per-session metadata stay byte-identical to pre-ranxianglei#5765 while off
  (zero footprint: no metadata key written while off — the storage-render
  goldens match master byte-for-byte).
- dispatch: enabled calls run the kernel ops exactly as before; a stray
  call while disabled gets a model-facing "not enabled on this bili proxy
  (compress.blockOps.enabled is not true)" receipt and never touches state.
- wire-contract goldens revert to the master shapes (the default surface
  is unchanged), and this commit supersedes the three earlier fix commits
  on this branch (test-pin updates + storage-render regen + payload
  retunes): with the default surface back to the master shape, those
  test files return to the master versions and the 27 red CI tests are
  green again without any calibration changes.
- docs: compress.blockOps in configuration.md/.zh-CN.md + website seed;
  kernel 0.0.194.

Verified: kernel 1656/1656; host targeted 201/201 (the 15 previously
red CI files + block-ops/snapshot/wire-contract), 348/348 + 63/63
surface/config invariant files; typecheck + build green.
…el 0.0.195)

Conflicts: kernel version files (master 0.0.193 from the ranxianglei#6132 fold-point
removal + this branch's 0.0.194 -> 0.0.195) and one relay.ts import line
(master dropped relocateCompressCarriers; this branch adds
buildBlockOpsSystemPrompt — the resolved import carries only the latter;
no fold-point call sites survive in the merged relay).

Re-verified on the merged tree: kernel 1631/1631, host targeted 201/201 +
surface/config 374/374, typecheck + build green.
… mcp-note split

unused-exports guard: MCP_DIALECT_NOTE_BLOCK_OPS and mcpDialectNoteText
are consumed only inside compress-tool.ts (withMcpDialectNote) — make them
module-local.
… gate

- lock-sync: the second master merge auto-merged the root package-lock
  with the pre-merge kernel version (0.0.193); regenerate it for 0.0.195
  (npm install --package-lock-only).
- lane-guard: src/server/side-request.ts (BILI_TOOL_NAMES += the kernel-
  fixed block-op names) is side-lane logic — pin it in the dedicated
  side-lane suite: tests/side-request-isolation.test.ts now asserts the
  bili-owned set covers ACP_TOOL_NAMES + block ops + absorb/rule/retrieve.
…ive-compact-known-conflict

[PR] thirdparty-scan: flag dsh-live-compact as a known conflict (ranxianglei#5984)
…emon

[PR] feat: compress-block retirement (acp_retire/acp_restore) + external summary rewrite (acp_rewrite) + export ledger
…r-digest-parity

# Conflicts:
#	src/state-hydration.ts
@ranxianglei ranxianglei changed the title fix: merge current master into ref-floor parity branch [PR] fix: merge current master into ref-floor parity branch Nov 11, 2026
@ranxianglei

Copy link
Copy Markdown
Owner

🤖 Powered by ework · qwen3.8-27b

(ai: vllm-qwen/qwen3.8-27b)

@Maziar123 — I mapped the topology and independently verified the conflict resolution against the live refs before touching anything. One hard stop on the final step:

I can't perform the merge. AGENTS.md makes PR merges human-only ("NEVER merge PRs — Human-only"), and that covers the git-equivalent too (checking out the branch, merging master, pushing to the owner's repo / rewriting a contribution branch). So the "Maintainer action" is yours to run. Everything up to it I've validated below.

Verified against the live merged head 70497b4c — all four claims hold:

  • Top commit is literally Merge remote-tracking branch 'origin/master' into 2026-11-03_ref-floor-digest-parity, sitting on current master (dc041dd1ab). It genuinely is "master merged in."
  • The collision was real: master defines COMPRESSION_STATE_FIELDS as a probe-based IIFE including timestampsHighWaterRefSeq (src/state-hydration.ts:107,109); [PR] Fix permanent unified-shadow DIGEST DRIFT on ref-floored sessions (Codex Desktop): replay leg minted a phantom byRef entry #5597 had its own array-literal version plus REF_FLOOR_RAW_ID.
  • In the merged tree both survive — probe-based COMPRESSION_STATE_FIELDS + timestampsHighWaterRefSeq (lines 50–53, 107–109) and REF_FLOOR_RAW_ID = "bili:ref-floor" (line 123). Union resolution, not a drop.
  • Diff of the merged head vs current master = exactly 9 files: src/session.ts, src/state-hydration.ts, src/storage/{ingest,mysql,shadow,store}.ts, tests/{issue5589-orphan-journal-ghost, issue5589-ref-floor-parity, storage-orphan-representable}.test.ts — matches your "same nine source and test files."

I did not re-run build/typecheck/the focused suites locally (you reported them green; per local-testing discipline the full-suite verdict belongs to CI).

Topology (so it lands where you expect):

To finish (pick one):

Verdict: technically sound — resolution preserves both fixes, footprint is the expected 9 files, merge is a clean FF. Ready for you to merge; I've left the merge itself to you per the human-only rule.

@ranxianglei

Copy link
Copy Markdown
Owner

🤖 Powered by ework · qwen3.8-27b

[bot] 🏷 (ai: vllm-qwen/qwen3.8-27b)

Closing note — full verification is on the floor above; here's the bottom line:

@Maziar123

Copy link
Copy Markdown
Contributor Author

Closing this PR because it is no longer needed.

#5597 already has master merged into it (9a558e7f, 2026-11-12). Its head keeps both changes this PR carried: the probe-based COMPRESSION_STATE_FIELDS with timestampsHighWaterRefSeq, and REF_FLOOR_RAW_ID. Its diff against master is the same 9 files.

Also, for the record: I have no write access to this repository, so I could not merge into the 2026-11-03_ref-floor-digest-parity branch here.

@Maziar123 Maziar123 closed this Nov 15, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.