@sema-agent/client-core 0.82.2 → 0.82.4

package/CHANGELOG.md

@@ -49,6 +49,19 @@
 > 挡住 ⇒ 本批把它机械化——④a0 对 `pending` 行**要求段头已是日期形**(`(未发布)` 直接红),阶段一
 > commit 漏转在发布前就红,不再靠人记。
 
+## 0.82.4(2026-09-24)
+
+### Added
+- **park 行审批卡新增过渡位 `irreversibleParkGate`**(CC-180):park 行重新出卡时,待决行的门类恰为 `irreversible_ask`,则卡入参带 `irreversibleParkGate: true`。这一位是门类事实,不是「必须真人过目」的无条件保证:引擎把首判即要求真人过目的 ask 在 park 时升成这一门类;安全轴收紧的门也落在这里(静态不可逆的工具、外发类工具,以及被 shell 闸收紧的 Bash —— 开了 shell 闸的部署上,大部分 Bash park 都会带这一位)。已知漏口(上游缺陷,已报):只在存储形变后的复判里才提出真人过目要求的 ask,引擎在 park 时丢掉这一要求、停成普通门,两位都不带。宿主可据此在 park 行卡上不给会话级放行,覆盖面宽、方向安全。在待决列表带出 `requiresRealApproval` 之前,这是 park 行上能读到的唯一信号。它不冒充 `requiresRealApproval`(两位各说各的事实),活卡那一路不铸。接入文档 **§96 S-2**。
+
+### Changed
+- **park 行上的审批卡在待决行带「必须真人过目」位时透传它**(CC-180):工具审批经 park 行重新出卡时,若待决行顶层带 `requiresRealApproval: true`(只认严格 `true`;嵌套在别处的同名键不读),卡入参同样带,与活卡那一路一致。🔴 生效前提:待决列表(`GET /v1/approvals`)要带出这一位 —— 今天的待决列表还不带,所以这一版在现有部署上**零行为变化**,park 行卡上这一位仍恒缺席;待决列表带出之后即生效,宿主零改动。接入文档 **§96 S-1**。
+
+## 0.82.3(2026-09-24)
+
+### Added
+- **自愈提示可以带「保住这段对话、把卡找回来」的出路**(CC-177):卡重开不了的两句(计划审阅卡 / 审批与提问卡)此前只给两条路 —— `/clear`(开新会话,这段对话就丢了)和手工向引擎发决断请求;而真能救回的往往是退出后续开本会话。`ActiveRunSelfHealCopy` 新增 `keepSessionWayOut`:宿主按此刻走不走得通,返回一个动作半句(小写起头、不带句末标点,例如退出后按会话 id 续开),走不通返回 `null`。包把它排在「放弃」那一路之前,两条路都给;句子说的是「续开后卡若还在等就会回来」,不承诺一定回来。以下情形不给、也不调用宿主函数:注入件、「决断还在路上」、宿主用 `rowFor` 整行覆写了这一结局,以及这道门的待决项已被证实不在引擎上 —— 这一形的结局新增机读位 `pendingRowGone: true`(续开也找不回)。这个证明只在呈卡**之前**算数:清理卡在屏上期间 run 可能转去等另一道门,所以呈卡之后的收口不带这一位,除非之后复读确认 run 已经结束。宿主函数抛错或返回空串按走不通。另新增 `engineDecidePath: false`,关掉交互面各行「直接在引擎上发决断」那半句;取消入口不受影响,headless 面也不受影响。两位都不给时,措辞与 0.82.2 逐字节相同。接入文档 **§95 S-1**。
+
 ## 0.82.2(2026-09-24)
 
 ### Added

package/README.md

@@ -35,7 +35,7 @@ Renamed from **`@sema-agent/wire-cc-adapter`** (0.1.x, deprecated — see *Migra
 
 ## Scope
 
-**Version:** 0.82.2
+**Version:** 0.82.4
 
 - **Today** — the adapter seam, the whole `adapt()` pipeline (all 14 A-layer arms plus the
   B/D/E tool-card layers), the notification/caps/model families, the adapter kernel (stream driver
@@ -361,7 +361,7 @@ guard still cross-checks the table by name).
 | `scripts/run-client-core-singleton-test.mjs` | Module-level singletons ⇄ `docs/refactor/p1-scan/singleton-manifest.json`, **both directions**: an unregistered singleton is red (registering it forces someone to answer "what if this got duplicated"), a stale entry is red, and the `dupRisk: high` count only goes down |
 | `scripts/run-catalog-loader-gates-test.mjs` | The model-catalog candidate chain (`loadCatalogWithSources`) and the provider device-code seam: offline ⇒ `bundled` with an honest `online.reason`, a good source ⇒ `online` plus a cache write, a second offline run ⇒ `cacheHit`; the three hostile source shapes (malformed JSON, `schemaVersion: 99`, off-domain `http`) each fall through to the bundled table, and an off-allowlist target is **never dialled** — including a `302` to another host, proven by a real loopback server's hit counter staying at zero; a one-byte edit to `catalog.sha256` drops that source while an unavailable sidecar only warns; and the device-code poller's `pending → ok` / `expired` arms run against a real loopback HTTP server with an injected clock |
 | `scripts/run-abortable-sleep-test.mjs` | The shared `abortableSleep(ms, signal)` leaf (consumed by `workflowClient.ts` and `agentSession/backgroundView.ts`'s poll backoff): normal timeout resolution, immediate wake-up on `abort` mid-wait, `clearTimeout` really firing on that path, and a post-resolve late abort staying a no-op |
-| `scripts/run-durable-card-display-keys-test.mjs` | The durable approval row's two display keys survive the row→card recast in `surfaceFsApprovalAndDecide`: `governanceForced` stamps on strict `true` only (absence is "no evidence", never `false`), `ruleSuggestions` passes through the same shape-narrowing reader as the live-frame leg and lands on the **read-only** card key — plus a standing pin that the durable leg never stamps the redeemable `ruleSuggestions` card position (the `/decide` body has no rule slot; offering a "don't ask again" option there would be an affordance nothing can honour), and a section for the parked twin of the classifier-unavailable fact: the upstream declares that key on the parked action itself, verbatim and under the same name as the synchronous ask, so this leg reads it rather than guessing a carrier name the way the deliberately unprojected keys must. The guard drives both legs with the same cause and asserts the card ends up byte-identical either way — the observable consequence of one reader serving two key paths, and the thing that silently diverges the day someone writes a second copy. Its own reach is printed rather than implied: what is proven is the package-boundary promise "on the row ⇒ on the card", not that today's engine flattens that key onto the pending row. A further section covers the two display facts the recast had been dropping for far longer. One of them the row has carried all along under a DIFFERENT NAME than the live frame uses — the frame puts it at the top level, the row nests it under the risk descriptor — and that difference in name is exactly why it went unnoticed; unlike the keys this leg deliberately refuses to project, its carrier is witnessed in the engine's own artefact rather than guessed. Neither is decoration: the shell's stand-aside arm reads them, so a call that matched a remembered allow rule which could NOT silence it looked like an ordinary ask on the durable path and was auto-approved with no card at all. Both land on the SAME card slot the live leg uses (one shape for the ends), verbatim bytes, present only when non-blank, never folded into an empty string — and the guard pins the discipline in both directions, including that a top-level key the upstream row does not actually have must still not grow this position |
+| `scripts/run-durable-card-display-keys-test.mjs` | The durable approval row's two display keys survive the row→card recast in `surfaceFsApprovalAndDecide`: `governanceForced` stamps on strict `true` only (absence is "no evidence", never `false`), `ruleSuggestions` passes through the same shape-narrowing reader as the live-frame leg and lands on the **read-only** card key — plus a standing pin that the durable leg never stamps the redeemable `ruleSuggestions` card position (the `/decide` body has no rule slot; offering a "don't ask again" option there would be an affordance nothing can honour), and a section for the parked twin of the classifier-unavailable fact: the upstream declares that key on the parked action itself, verbatim and under the same name as the synchronous ask, so this leg reads it rather than guessing a carrier name the way the deliberately unprojected keys must. The guard drives both legs with the same cause and asserts the card ends up byte-identical either way — the observable consequence of one reader serving two key paths, and the thing that silently diverges the day someone writes a second copy. Its own reach is printed rather than implied: what is proven is the package-boundary promise "on the row ⇒ on the card", not that today's engine flattens that key onto the pending row. A further section covers the two display facts the recast had been dropping for far longer. One of them the row has carried all along under a DIFFERENT NAME than the live frame uses — the frame puts it at the top level, the row nests it under the risk descriptor — and that difference in name is exactly why it went unnoticed; unlike the keys this leg deliberately refuses to project, its carrier is witnessed in the engine's own artefact rather than guessed. Neither is decoration: the shell's stand-aside arm reads them, so a call that matched a remembered allow rule which could NOT silence it looked like an ordinary ask on the durable path and was auto-approved with no card at all. Both land on the SAME card slot the live leg uses (one shape for the ends), verbatim bytes, present only when non-blank, never folded into an empty string — and the guard pins the discipline in both directions, including that a top-level key the upstream row does not actually have must still not grow this position. The security-class approval bit (`requiresRealApproval`) rides a parked row's card when the row carries it at top level, on strict `true` only, while look-alike nested carriers are ignored; this is pinned with a constructed row, because today's pending list does not carry the bit yet. A second, separate bit (`irreversibleParkGate`) marks a parked card whose row sits on the irreversible-ask gate kind — a gate-kind fact that covers asks the engine flagged for real approval at the first decision plus safety-tightened gates, with a known engine gap for approval demands raised only on a storage recheck — on the park path only, on the exact gate word only, and never in place of the real bit. |
 | `scripts/run-session-memory-status-test.mjs` | The session **memory-status** read face (S-53): the two judgements three clients would otherwise each get wrong. First, *same status, different code* — this route's 404 carries two unrelated meanings (`not_found.session` = unknown or non-owned session; `not_found.route` = a pre-7.53 server that has no such route at all), so dispatching on the **status** would report "your deployment lacks this surface" as "your session does not exist". The verdict is anchored on `errorCode`, the two 404s are pinned to **different** verdicts, and — the load-bearing negative control — a 404 carrying **no** code falls to `failed` rather than guessing either way, since a wrong guess in either direction is a false statement a user would act on. 501 is allowed a codeless fallback because both of its arms mean the same thing here, and `capability.*` stays split from `feature.*` because those two share a status while their dispositions are opposite. Second, *absence means something different per key*: `optOutSource` and `lastCaptureAt` are legitimately absent on a **healthy** session (a zero-history session really is `{captureOptedOut:false, committedCount:0, foldedCount:0}` with no degradation at all), so reading absence as "off/none/0" asserts something unprovable. Two combined readers are pinned: capture opt-out is read from **both** its keys (a record-store fault yields `indeterminate`, never `active` — the difference between "your conversation is being remembered" and "nobody knows"), and last-capture is a **three-state** read whose discriminator is the *other* key, because `lastCaptureAt`'s absence alone covers both "ledger unreadable" and "genuinely no contributions" and therefore decides nothing; the two shapes are pinned to different verdicts so a single-key read turns red. The thin wrapper is the only IO: it never throws, drops malformed keys to absence rather than trusting them (an unreadable value must answer "don't know", never render as truth), refuses to spend a request on an empty `sessionId`, and passes `signal` through untouched |
 | `scripts/run-crash-converged-projection-test.mjs` | The `crashConverged` read face on `GET /v1/approvals` (L-38): what the *previous life* of a crashed local engine left behind, projected for every client. Three judgements are pinned. First, **absence is not an empty list** — a missing key (an older server, deps not present, or a carrier that is not an array at all) returns `undefined`, and the client renders nothing; an empty array returns a present zero-count object, which is the server actually saying "none". Folding the first into `{total:0}` would have the client assert "nothing was left behind" on a surface a person uses to decide whether it is safe to re-run something — the worst possible direction for a false statement — so the two cases are pinned to different **return shapes** and a test asserts the two verdicts are unequal. Second, bucketing is a **four-term conjunction**: `orphanState === 'pending'` *and* `resumeSafe === true` *and* both approval-evidence keys (`originalDecision`, `decidedAtMs`) absent. A fifth term rejects any row carrying an **accessor**, and accessors are never invoked at all — reading one means synchronously running someone else's code, and `catch` catches throwing, not *never returning*, so a looping getter would pin the startup thread forever (the row cap does nothing against that shape). The same rule covers the three untrusted reads outside the row as well — the envelope's `crashConverged` key, the carrier's `length`, and every numeric index are read as own property *descriptors* and only data descriptors are used, so accessors and prototype entries read as absent and are never invoked. Such a key is treated as absent: if it was a required field the row is counted as dropped, if it was optional or additive the row survives without it. That also closes the ordering attack, since spreading runs getters in property order and an earlier one could `delete` the approval evidence before it is ever copied (measured before the fix: such a row reached the resume-safe bucket), and the check therefore moves ahead of the read, onto the property descriptors — from which the snapshot is then built directly, because checking descriptors and *then* spreading is two independent observations of the same row, and a non-throwing proxy can make the two `ownKeys` calls disagree (first showing `originalDecision: 'approve'` so the row reads as plain data, then omitting that configurable key so the snapshot loses the evidence; measured before the fix: the dangerous row reached the resume-safe bucket after exactly two enumerations, and after it, one). Keys are written with `Object.defineProperty` rather than plain assignment, because `'__proto__'` is a legal own enumerable key and `o['__proto__'] = x` does not store a value — it calls the prototype setter, letting a row whose own properties are all plain data (so the accessor gate never fires) inject a prototype whose `sessionId` getter deletes the approval evidence from the snapshot during validation; `defineProperty` fires no setter, so the key survives as ordinary additive data and the snapshot keeps `Object.prototype`. A row that simply arrives with a custom prototype is treated the same way, since the snapshot only enumerates own properties: approval evidence sitting on the prototype would never reach it, and a perfectly ordinary object with no proxy and no accessors could otherwise be called safe to re-run — real bodies come from `JSON.parse` and always carry `Object.prototype`, so nothing genuine trips it). Validation itself runs on a **null-prototype** dictionary and the bucketing verdict is carried out of that same pass rather than re-read from the delivered row, because every property lookup on an ordinary `{}` reaches `Object.prototype`: a polluted `sessionId` getter there would delete the approval evidence from the snapshot mid-validation and send the row to the safe bucket (measured before the fix). The row handed to the client is still an ordinary object — the null prototype is an implementation detail of the check, not of the value) — real JSON bodies are all data properties, so only a middle-layer-synthesised payload ever trips it, and it too lands in the human bucket rather than being dropped. The `decided` arm means the human had already approved and side effects may be half-landed, so it always goes to the human bucket, as does `resumeSafe === false` and — the last two terms — any row whose own fields contradict each other, since `pending` claims nothing ran while that evidence says somebody pressed approve. Deciding "not safe" costs one extra question (recoverable); deciding "safe" wrongly has somebody re-run work that already partly happened (not). A 2x2 truth table pins that exactly one cell is resume-safe, so reading either key alone turns red, and the contradictory rows are routed to the human bucket rather than dropped — they are real orphans, and the ones most worth showing. Third, unreadable rows are **dropped and counted**, never thrown and never passed through: the product is declared as `CrashConvergedRow`, so letting a row missing a required field — or carrying one of the wrong type — past would be a lie at the type level, and the closed literal discriminators (`decision` / `cause` / `orphanState`) decide family membership rather than being an open vocabulary. The measuring stick stops at the **type** floor, though: degenerate-but-well-typed values (`ts: NaN`, an empty `toolName`) are kept, because swallowing a real orphan over a decorative field is the worse direction, and the one deliberate exception is `approvalId`, which must be non-empty to be a row identity at all. `dropped` is kept separate from `total` so unreadable rows never inflate "N approvals were affected"; each row is a **one-shot snapshot** — every own enumerable key is read exactly once, and validation, bucketing and the handed-back value all read that same snapshot, so additive upstream keys survive while a **non-idempotent** getter (one that never throws, just answers differently on a second read) can no longer erase the approval evidence between the check and the bucketing (measured before the fix: such a row landed in the resume-safe bucket while its checked value was `"approve"`). Hostile carriers are counted rather than allowed to reject: **every** touch of the carrier is guarded — envelope property reads, `Array.isArray` itself (it throws on a revoked proxy), the `length` read, each indexed read and each row's property reads — and a traversal that dies halfway returns absence rather than a half-counted total. A row that cannot be read never takes the batch with it: its own shape check is inside its own guard, so one revoked-proxy row costs a `dropped` tick rather than collapsing the whole projection to absence — which a client would have read as "this deployment does not offer the surface". Traversal goes by **numeric index, never the carrier's own iterator protocol**, because `for...of` hands the carrier the question of which rows exist: an array carrying an overridden `Symbol.iterator` can yield nothing (measured before the fix: a real orphan became `{total:0}`, which a client reads as "the server said there are none") or swap a dangerous `decided` row for a safe-looking one (measured: `fake-safe` was returned in place of `real-danger`). Row count is capped at 100000 and the cap is checked **before** the walk: requiring only a non-negative integer `length` does not stop a proxy trap reporting a billion, and this surface runs on the startup / `--resume` path, where a synchronous spin freezes the thread (measured before the cap: twenty million rows took 18.3 seconds and twenty million index reads; a billion does not come back). The honest boundary is stated rather than overclaimed — a proxy can still lie in its `length` or index traps, which is the same thing as a host injecting a lying transport — and the widening of `ApprovalsResourceLike.list()` is proven **additive** by really running tsc over a legacy `{ pending }` mock *and* over the real `AgentClient` path — the projector takes `unknown` precisely because a parameter shaped as "an object with an optional `crashConverged`" is a TypeScript weak type that the installed SDK's own `list()` return shape shares no property with, which only a real-client compile would have caught — with a known-red control so a clean run means the checker spoke |
 | `scripts/run-self-orchestration-denial-test.mjs` | The three judgements behind a **denied self-orchestration request** (server 7.57.0), each of which all three clients would otherwise get wrong on their own. First, whether to retry at all is a **conjunction that may not be loosened**: HTTP 501 *and* an `errorCode` that is **exactly** `capability.self_orchestration_required`. That code shares its shape with every other `capability.*` 501, so dispatching on the prefix would drag "some other capability is not wired up" into the retry arm — those requests do not become acceptable once the two keys are gone, so the client would spend a request and then tell the user the wrong reason. Negative controls cover all four directions: a sibling `capability.*` code, a truncated or suffixed variant of the right one, a codeless 501 (it decides nothing, so it decides nothing — no guessing), and the right code under 500 / 400 / 503 or a string `"501"`. The classifier reads structurally rather than by `instanceof` (a host may inject its own transport; across realms or duplicate SDK instances an understandable error would read as unreadable), so a class instance, a bare `{status, errorCode}` literal and an error carrying those fields on its **prototype** all reach the same verdict — and a hostile proxy or a throwing getter yields `null` instead of throwing, because this classifier runs inside a `catch` block where anything it throws escapes the caller's own guard. Second, removing the intent is a **structural** operation, not wording: `selfOrchestration` sits at the top level while `ultracode` sits under `settings` — two different stamping legs — and a client hand-writing `delete` will miss the second one, which costs the user the same failure twice. The single stripper is pinned to touch exactly those two: other `settings` sub-keys and their values survive byte for byte, `deferTools` is left alone (pulling `Workflow` out would be a behaviour change, not a removal of intent), additive unknown keys survive at both levels, the input object is never mutated, `settings` is only dropped entirely when `ultracode` was really there and nothing else remains (an already-empty one is left as is), a non-object `settings` is not touched at all, an `ultracode` that only exists on the prototype does not count, and the whole thing is idempotent. The end-to-end leg runs a real `buildTaskRequest` product through it and asserts the stripped body still passes the registration gate key by key. Third, on the capabilities body, **absence is not "switched off"**: a pre-7.57 server has no `workflowsGate` key at all, so reading absence as "the engine says no" asserts something the server never said, and the mirror-image disease is folding an **unrecognised** `denial` into `null`, which would have the client render "nothing was denied" when the truth is "denied, for a reason I do not recognise". Five shapes are pinned — caps unreadable, gate absent, closed-set member, unknown value, accessor — with the unknown arm carrying the raw token (or an empty one when the value is not even a string) and never collapsing to `null`. All four untrusted reads go through own **data descriptors** only, and the guard pins the getter invocation count at zero, since `catch` catches throwing but not *never returning*; a descriptor trap that throws and a revoked proxy both yield honest absence rather than an exception — though *what* absence means differs by field, and the guard pins that split rather than a blanket rule: an accessor on `workflows`, `workflowsGate` or `engineCan` reads as absent, while an accessor on `denial` reads as `{unknown:''}`, because a key that is **not there** is the gate saying "nothing was denied" whereas a key that is there but cannot be read is "denied, and I could not read why" — folding the second into the first is exactly the false statement this face exists to prevent. Two further pins came out of an adversarial review. The exported retry list is **frozen at runtime**, not merely `as const`: the verdict hands out that same reference, so any consumer splicing it once would poison every later verdict in the process — the guard asserts `Object.isFrozen`, that four different mutation attempts leave it byte-identical, and that a verdict issued *after* those attempts still carries the original two entries. And the classifier reads `denial` only **after** both criteria have passed, since it is not a criterion but an extra field on the verdict: the guard pins the getter invocation count at zero for any error that does not match and at most one for an error that does. The scope line is drawn explicitly rather than overclaimed — "no getter ever runs" holds for `projectWorkflowsGate`, which reads **wire JSON** where every field is an own data property by definition, but not for the classifier, which reads a **thrown value** that may well be an SDK `APIError` class instance carrying `status` and `errorCode` on its prototype; insisting on own data descriptors there would report a perfectly readable error as unreadable, so that side promises only that it never throws. A final pin covers the **integration document's own worked example** rather than the library: the shipped SDK's `tasks.stream()` is an `async` generator, so calling it issues no request at all — the POST happens inside `streamRaw` on the first iteration, and a `try` wrapped around the `stream(...)` call itself can never catch the 501. A client following a submit-shaped recipe on the streaming leg would never run the classifier, and the whole strip-and-retry path would silently do nothing. The guard drives the **real** `TasksResource` against a fake transport, offline, and pins both halves: the synchronous leg is in flight the moment it is called, the streaming leg has issued zero requests after the call and raises on the first `next()` — and it does so through the **real** error path, with `openStream` returning an actual 501 `Response` that the SDK's own `errorFromResponse` turns into the typed error, pinning the `openStream`→`errorFrom` call order so a transport that stops minting `errorCode` cannot pass. The documented recipe is then **executed** rather than keyword-counted: exactly one retry, a second body that really lost both keys while every other setting survives byte for byte, the caller's own request object left untouched, one disclosure and only one, a second 501 propagating with the request count still at two, and — after the first 501 — an abort leaving the count at one with nothing disclosed. A last leg is type-level: `stripSelfOrchestrationIntent` carries an SDK `TaskRequest` overload, because the wide `Record<string, unknown>` form erases the caller's type and the document's "strip and resubmit" line would not compile without an unsafe cast; a real tsc run over a virtual file proves both the narrow and the wide path, with a known-red control — and it compiles the document's two recipes **verbatim**, extracted from the section itself, because a recipe that does not compile is a recipe that was never given: `{ transientOk: true, signal }` is a TS2379 under `exactOptionalPropertyTypes`, which no amount of prose review had caught. The last thing pinned is the one that would have been quietest of all: the SDK's `stream()` returns only on a `done` or `failed` frame, so a stream truncated mid-run — or yielding nothing at all — ends the `for await` just as normally as a completed one. The documented `runOnce` therefore tracks whether it ever saw a terminal frame and raises when it did not, the guard's success fixture emits a real terminal and asserts the handler received it, and a truncated-stream control asserts that shape is reported as a failure with no retry and nothing disclosed. That terminal-frame rule then needed one more turn of its own: the underlying reader returns *normally* when the signal is aborted, so the check as first written rewrote a user's cancellation into a generic stream fault — a client keying off `AbortError` to suppress the error would instead have shown a failure, or resubmitted. Cancellation is therefore checked first, a real-SDK case aborts from inside the handler and asserts the original `AbortError` survives with no retry and nothing disclosed, and the document is checked for that ordering. The harness runs the documented `handle` and `transcript.note` as real spies rather than pushing frames itself, the drive loop rethrows exactly as the document does, and the disclosure ledger is proven to be the caller's own array by a positive identity assertion — without which the cancellation leg's "nothing disclosed" would have been vacuously true. Each recipe is compiled **on its own**, with a preamble that declares only what a host supplies and injects no library symbol, since compiling them together let the second one borrow the first one's imports, and the preamble's own types are decoupled from what the recipes import so the "remove the imports and it must fail" control fails for the right reason — which is checked by attribution, not merely by redness. Ordering is the last thing to get right: the cancellation check must come before the truncation error but **both** must sit behind the terminal-frame test, because a cancellation that lands after the run already reported `done` would otherwise overwrite a real outcome — one that may have already had effects — with "cancelled", and a person reading that will run it again. Aborting from inside `handle(done)` and `handle(failed)` are both pinned to still report success, and the ordering assertion is anchored inside the streaming `runOnce` body rather than the section, since the section's first `throwIfAborted` belongs to the synchronous recipe and would have made a reversed streaming recipe pass — and that ordering check is now anchored on the TypeScript AST rather than on text, since a comment reproducing the two statements in the right order let a genuinely reversed body pass. One more timing fact had to be written into the recipe: a single SSE read buffers several frames and the SDK yields them back to back, so checking the signal only after the loop lets a cancelled run keep consuming the rest of the chunk — measured, an abort inside `handle(turn_start)` still swallowed the `done` that followed and reported success. The recipe therefore re-checks after every non-terminal frame. Finally, the behavioural matrix is no longer run against a copy of the recipe: both recipes are extracted from the document, transpiled, and **executed** with injected host objects, so the disclosure assertion really exercises the document's own `transcript.note(disclose(...))` line, and the synchronous leg gets the same full matrix the streaming one does |
@@ -370,11 +370,11 @@ guard still cross-checks the table by name).
 | `scripts/run-type-superset-ledger-test.mjs` | The type/wire **superset ledger** (`docs/type-superset.json`): positions this package adds on top of a CC-shaped contract, each carrying the evidence for what CC's own type surface does or does not have there. Completeness is deliberately uneven and the ledger says so. The `_sema_*` private-key class is checked in **both** directions (a key in the source that never entered the ledger is red, naming key and file; a ledger row whose key left the source is red) — but only for keys written as literals, which is the convention the ledger mandates. A key assembled by string arithmetic is beyond what any static rule can enumerate, so the guard fails closed on every shape it *can* decide (a bare `_sema_` prefix is red wherever it appears, save one pinned guard site) and leaves the rest as a convention violation for review to catch, rather than claiming a completeness it does not have. The two hand-surveyed classes are only checked for coordinate and evidence integrity, never discovered. Both directions read the source through the **TypeScript AST**, not a text scan, and they read two different sets out of it. A *key site* is an identifier, or a string whose whole value is the key — so `'_sema_decision-v2'` is carried whole rather than truncated at the first non-identifier character into some *other* key that happens to be registered. A *mention* is the key appearing inside a longer string, which is prose, not usage. The staleness direction counts key sites only: a comment or a doc sentence left behind after the last real mint site is deleted must not keep the row alive (mutation-proven — with both the comment and the prose string untouched, removing the one real site turns the guard red). And because a prefix can be concatenated or interpolated into a key no static set will ever see, the bare `_sema_` literal is refused outright rather than traced: every occurrence is red except the single inline `startsWith` guard the sanitizer needs, because the set of expressions a bare prefix can travel through on its way to a concatenation is open-ended and enumerating it is always one form behind. Every row's `host` must still resolve, with the key being a real **member of that declaration** rather than a string occurring somewhere in the same file — `governanceForced`/`delegation` each live on two different shapes in one file, and a member commented out is a member deleted, which a text-shaped check happily reads as still present. And the direction worth the most: each machine-form `ccAbsenceEvidence` is re-derived from the row's own `key` — the ledger's recorded string must match that derivation verbatim, since a row quietly witnessing `\bnever_present\b` is green forever while watching nothing (mutation-proven: the same edit passes the unbound form and is caught by the bound one) — and the check runs against the names the installed `@sema-agent/agent-types` `.d.ts` set actually declares, parsed with the TypeScript AST rather than grepped, so a name CC merely mentions in a comment cannot force the row into the manual escape hatch and thereby retire the very witness that was supposed to fire the day CC declares that name for real. That escape hatch is gated by an allowlist living **in the guard**, not the ledger, so claiming it costs a reviewed diff. Missing material never reads as a pass, and the verdict splits by *why* it is missing: no TypeScript parser skips the suite before it starts; a missing `agent-types` still runs and prints the first three directions, then exits **1** when `package.json` declares the mirror but it is not installed — a broken install must not retire the repository's only "the day CC declares this name" alarm, and reporting it as a skip would leave "never evaluated" and "evaluated, no drift" indistinguishable to the runner — and exits 3 only when nothing declares the mirror at all, which is the one case where the direction genuinely does not apply. Either way a run that evaluated no witness is never counted as one that did. When the mirror *is* present its **installed version** is witnessed too (the two declared floors must agree with each other and the installed copy must meet them), since four preflight probes are satisfied by an arbitrarily stale mirror — they prove the extractor speaks, not that it is current. Every direction carries a positive control — known-present CC symbols, a comment-only sample proving the extractor distinguishes declaration from mention, and synthetic corpora fed through the **same** discriminator function the real verdict uses, so a verdict quietly rewritten to return nothing takes its own control down with it |
 | `scripts/run-rules-side-test.mjs` | The persisted-permission-rules lane's shared decision half. The two capability bits are checked as **two independent gates** — a worker can honestly advertise the rules lane while predating the revoke routes, and that shape must *hide* the governance surface rather than render a dead entry. Failure classification is by **disposition, not cause**: the two 404s (route missing vs. dead ticket) never share a bucket, a 503 `rule_import_retry` means *the ticket is still alive* (the opposite handling of a dead one), and a stale-cursor 400 drops the cursor and re-lists from the top exactly once — never resuming a stale keyset, never surfacing a partial governance list, and never paging past the hard cap. The persist-ack reader is **merged into** `readToolApprovalRespondAck`: the three-state verdict (`persisted` / `refused` / `unknown`) is derived only from an ack that passed the package's structural narrowing, and a half-shaped object such as `{rulePersisted: true}` with no `delivery` reads as `unknown` — the pre-merge shell read would have said `persisted`, which is precisely the double-ledger drift this file closes, so that case is pinned in reverse. The local-allow-rule skeleton pins all five narrowings (whole-tool, tool-name match, literal anchor with the escaped-star counter-example, bare interpreter prefix consulted only for Bash, and the canonical dangerous-pattern overlay) **with their refusal strings byte-for-byte** — the cli's 128-assertion suite anchors the same strings, so a one-character edit here changes observable behaviour on three clients — and asserts the parse is a pure function of its input, because the same call backs both "render the option" and "resolve the selected value" |
 | `scripts/run-park-decision-layer-test.mjs` | The decision layer behind the "stuck behind a card" family, shared by every client. A pending row that is **not in the queue** is three states, not one: a bounded, interruptible re-probe loop distinguishes *a decidable row*, *not born yet* (no positive evidence that anything settled — an empty queue proves nothing) and *settled elsewhere*, always probes at least once so a zero budget keeps the pre-fix semantics verbatim, cuts a hung read face off at the window rather than only noticing afterwards, and reports the honest failure when the window is spent instead of inventing a decision. The decision-note reader is likewise three-state: an explicit `noteRecorded: false` outranks an echoed note body, absence renders **no line at all**, and untrusted note text is flattened and bounded before it ever reaches a renderer. Row routing anchors on the deciding quantity — a row carrying `gateKind: "human"` with `toolName: "Write"` is a tool gate, because `human` is the engine's *generic* "someone must decide", not a synonym for a question — and the queue scan refuses to surface a row it cannot positively prove belongs to this session. A chain that fails after the row vanished is split by whether a card was ever presented: decided-elsewhere, or not-its-turn-yet. A row-level single-flight makes "at most one card per pending item" structural rather than incidental. The resume three-way card pins the option **order** (the zero-effect choice sits at index 0, because the frame carries no default-focus field and a stray Enter must not attach or cancel), renders only options the wired verbs can honour, collapses every ambiguous answer to zero action, omits the liveness line entirely when the engine gave no evidence, and — when there is no card lane at all — prints three real routes and exits on a dedicated code rather than reporting success |
-| `scripts/run-selfheal-reopen-test.mjs` | The 409 active-run self-heal decision chain: `governanceForced` narrows on strict `true` only; triage prefers the wire's `pendingGate.kind` and falls back to the status table (an off-table kind is never guessed into a card arm — hands-off plus the honest wording); a first-sight card makes zero closed/reopened claims and a host presentation receipt of `presented: false` demotes the outcome to reopen-failed; park-row ownership is a fail-closed positive proof (own-run ledger or session id — unprovable is not owned); the three gate-identity key literals live in exactly one mint (`hitl/gateIdentity.ts`, AST string-token scan); the armed-gate presentation ledger is per-session; and the `plan_review` reopen arm shares the arm arm's card body, three-state verdict and delivery pipe, consuming the presentation history once a decision is delivered. The same chain also carries the `running` three-way card: both plan-family gate kinds route to the plan arm and all four ask-family kinds to the ask arm (an off-table kind still never gets guessed into either); the card is offered only for verbs that can actually be honoured and a missing presenter means zero action rather than a silent cancel; a steer is sent **exactly once** with its three delivery outcomes worded apart (a `queued` receipt is the wire correcting the triage input, so the named park word decides which card gets reopened, and an unrecognised park word drives neither arm), and a steer failure is split into *provably not delivered* (4xx) and *delivery unknown*, because telling a user to resend a non-idempotent instruction that may already have landed is how duplicates get made. After a user-chosen cancel, "the session is free" is asserted only from a whitelist of terminal states — park states hold the claim, an unrecognised state word is not a release, a failed read is *unknown* rather than a release, and only a 404 counts as one — and the honest timeout line quotes how long it really waited |
+| `scripts/run-selfheal-reopen-test.mjs` | The 409 active-run self-heal decision chain: `governanceForced` narrows on strict `true` only; triage prefers the wire's `pendingGate.kind` and falls back to the status table (an off-table kind is never guessed into a card arm — hands-off plus the honest wording); a first-sight card makes zero closed/reopened claims and a host presentation receipt of `presented: false` demotes the outcome to reopen-failed; park-row ownership is a fail-closed positive proof (own-run ledger or session id — unprovable is not owned); the three gate-identity key literals live in exactly one mint (`hitl/gateIdentity.ts`, AST string-token scan); the armed-gate presentation ledger is per-session; and the `plan_review` reopen arm shares the arm arm's card body, three-state verdict and delivery pipe, consuming the presentation history once a decision is delivered. The same chain also carries the `running` three-way card: both plan-family gate kinds route to the plan arm and all four ask-family kinds to the ask arm (an off-table kind still never gets guessed into either); the card is offered only for verbs that can actually be honoured and a missing presenter means zero action rather than a silent cancel; a steer is sent **exactly once** with its three delivery outcomes worded apart (a `queued` receipt is the wire correcting the triage input, so the named park word decides which card gets reopened, and an unrecognised park word drives neither arm), and a steer failure is split into *provably not delivered* (4xx) and *delivery unknown*, because telling a user to resend a non-idempotent instruction that may already have landed is how duplicates get made. After a user-chosen cancel, "the session is free" is asserted only from a whitelist of terminal states — park states hold the claim, an unrecognised state word is not a release, a failed read is *unknown* rather than a release, and only a 404 counts as one — and the honest timeout line quotes how long it really waited. The two "card could not be reopened" rows can carry a host-declared way to keep the conversation, which says the card comes back on resume only if it is still waiting: it is placed before the route that abandons it, never offered for an injected submission, while a decision is still on its way, or once the pending approval has been proven gone (the outcome then carries a flag saying so; the proof only counts before the cleanup card is shown, so a fallback after the card carries no flag unless a fresh read finds the run finished, and a recheck that finds the approval back clears it), the host function is not even called in those cases, and it is treated as unavailable when it throws or returns an empty value; a host can also switch off the engine decide route on the interactive rows while the cancel route stays, and with neither given all four rows are pinned byte-for-byte to the text the previous release produced. |
 | `scripts/run-terminal-identity-copy-test.mjs` | Terminal-state **identity**, in both lanes where a stop gets a name. A run stopped by this deployment's own governance knobs — the open-set `limits.*` family, `output.invalid`, and the `blocked` contract terminal a ReportBlocked agent produces — is not a provider failure, and labelling it `API Error:` sends the reader to check the network, the key and the quota when the handle is the `--max-turns` they passed themselves. Those terminals now render a neutral row; the reverse direction is guarded just as hard, because asserting "this is *not* an API error" on a code the package does not recognise is the same misfiling pointed the other way — a real `gateway HTTP 502`, a `conflict.session_active_run` and any unknown code all keep the `API Error:` prefix, and the row keeps its `isApiErrorMessage` class flag so brief-mode visibility filtering does not silently drop it. The second half is who the rejected submission belonged to: the self-heal copy told every caller "Your message was NOT sent … send it again", which is three separate untruths for a system injection (a plan-review outcome, a cron wake-up, a task notification) — not the user's message, and not re-sendable, since a host queue marks those non-editable and non-recallable. The injected form says so instead, and the one sentence that promises re-delivery is pinned to the single disposition that earns it: `selfHealSubmissionDisposition` is the same function the host consults before putting the item back on its queue, so the promise and the behaviour cannot drift apart, and the arms where no card could be surfaced state plainly that nothing was delivered and nothing will retry. Since 0.72.6 the same gate pins the **follow intent** after a steer (): a message handed to a live run only pays off if someone tails that run's own event stream, so `steerFollowIntent` decides from the delivery word whether to tail now, after the pending decision, or only after a wake — and the "watch that run" sentence turns into a factual "sema is following that run" **only** when the host declares it attached that tail, so a shell that did not wire it can never claim it did |
 | `scripts/run-additive-key-passthrough-test.mjs` | The one disease shape behind two legs: a **closed whitelist / flattening arm** dropping a fact that is already on the wire, while both sides of the seam look correct. (1) The `task_progress` projection carries a registered **key ledger** — a frame populated with every key the service really projects is pushed through the shipped `eventToSdkMessage`, and the set of wire keys that survive must equal the registered pass-through list **name for name in both directions**, so quietly forwarding one more key is as red as quietly dropping one. `model` (the child run's model id, minted by core as `prepared.model.id` and projected by the server since 7.52.1) is the key this batch adds, with the same conditional the server itself applies: a non-empty string or no key at all — an empty string is neither a model id nor "unknown". The ledger is also checked against the fenced list in `docs/INTEGRATION-CLIENTS.md` §3d, so a doc that still says seven keys while the code forwards eight is red rather than merely stale. (2) The decide-failure arms carry the server's S-02 `currentPending` pointer key from a 409 `approval_stale` refusal onto the outcome the host reads. The reader is structural rather than `instanceof`, because the client is host-injected and the class identity is not this package's to assume; a half triple never mints (half a pointer cannot relocate anything), an empty string is not presence, and `checkpointToken` never transits. Both the allow and the deny leg are driven end to end through the real durable approval path — as is the accept-session leg, where a refusal carrying the pointer key must now re-raise instead of silently re-sending the human's answer for the **old** card as a plain approve (one decide call, pointer preserved), while a legacy 400 still falls back exactly as before — and all three flattening points must call the one shared reader — the same-shape residue check that makes "fixed one arm and left the twin" red instead of invisible. (3) The same disease growing on the REQUEST side: the `.mcp.json` → server-spec projection rebuilds each server key by key, and the settings schema deliberately leaves some keys parse-transparent — whatever JSON the file carries reaches the engine untouched, because validating them where the whole domain parses all-or-nothing would let one bad declaration take every server down silently. The whitelist had no row for the newest of them, so an operator's per-tool declarations — the ones the write fence reads — were stripped at the package boundary while both sides looked correct. The criterion is not "is that key handled" but the transparent-key table read out of the INSTALLED schema at runtime, reconciled name-for-name against this leg's ledger, so the day upstream adds a third one this turns red and forces an explicit decision. Behaviour is pinned on both transports, by object identity rather than deep equality (a rebuild would be a second judge), and malformed values must transit UNCHANGED rather than be refused here — the engine refuses them loudly and names the server, whereas a package-side judge can only swallow a declared protection quietly. Absence still mints no key, unknown keys still never reach the wire (the fix is the dropped key, not the gate), and the one transparent key this leg deliberately does not forward is a ledger entry with its own exit condition: it belongs to the deployment plane, and the day the request-plane type declares it the entry's premise is gone and the gate says so |
 | `scripts/run-esc-halt-plan-test.mjs` | The Esc stop decision every client shares: fire the **turn-level** halt first, and escalate to a **run-level** cancel in exactly two cases — the engine itself answered with a 409 from the closed code set (it is saying "there is no in-flight turn here; use cancel for a run-level stop"), or that shot came back with no verdict at all *and* the shell can independently prove a permission card was on screen. Everything else does not escalate. The asymmetry is the whole point and every negative control guards the same direction — deciding *not* to escalate costs the user one more choice on a busy-session card (recoverable), deciding to escalate wrongly tears down a run that was alive and takes every in-flight tool with it (not). So: the closed code set is a **frozen** value, not a `ReadonlySet` — type-level immutability does not stop a consumer's `.add()`, and the guard proves it by really trying to mutate the exported value and then checking the verdict did not drift; the escalation gate is the **conjunction** of that closed set and the 409 status, since honouring the code alone lets a 500 that merely quotes it drive a destructive call; `interrupt.not_held` and `steering.not_running` are deliberately outside the set (the first means *this replica* has no live face — the run may be perfectly alive on another); an unreadable code falls to the no-escalation side; a `parked` flag never overrides a verdict the engine did give, and only strict `true` counts when it did not. The first shot is unconditional by construction — it does not consult `parked`, because the 409 it earns is exactly the verdict the gate wants — and the verdict itself is a closed machine-readable reason word, not display copy. A third escalating case was added once tearing the stream stopped reaping the run: with detach armed, a shot that never lands leaves the run going all the way to the end of the turn, so the Esc the user pressed has no effect at all and nothing on screen says so — the old behaviour had a silent backstop (tearing the stream ended the run) and that backstop is gone. The new fact is held to the same three disciplines as `parked`: it is read only where the engine gave no verdict, it is judged **after** `parked` so an existing host's reason word does not change under it, and only strict `true` counts. Absence is proven to be a no-op rather than asserted — the guard carries its own reference implementation of the previous version's table, runs the full grid through both, requires zero divergence when the new field is omitted, and first shows the comparison really does report a difference on the one cell where the two versions are meant to differ |
-| `scripts/run-peer-frame-projection-test.mjs` | The three engine-injected lanes design/385 puts on the **one** `task_notification` carrier, which are not the same kind of thing at all: a delegated child's uplink (`agentMessage`), another session's message drained from this session's own box (`crossSessionMessage`), and a receipt about one of *this* session's own outbound messages (`crossSessionNotice`). The engine renders none of them inside a `<task-notification>` shell, so a client that projects them as the generic completion card shows "background task finished" while the model read a colleague's sentence — two faces describing different events. The discriminator is pinned to the **typed carrier being present**, never to the `summary` text: those carriers can only be minted by the engine's injection legs (the external `notify()` input is a strict subset of the payload and can wear none of them), while `summary` is filled by every notification there is — so anchoring on text would let any background task impersonate a colleague's message by writing `<agent-message from="…">` into its own summary, and a positive control asserts exactly that payload still projects as the generic card. Fail-closed has two tiers rather than one: a broken **required** field (empty `from`, a non-string `body`, a notice `kind` outside the closed set) returns absence so the caller falls back to the generic card — an honest downgrade where the user still sees the notification — while a broken **optional** field drops only itself, because losing an attribution note and losing a colleague's whole message are not the same magnitude. The provenance side record is **required and must agree on four points** (`kind` matches the lane; `from`/`taskId`/`seq` are present and equal the carrier/payload — each equality is anchored on a core mint site and pinned by the cli wire-anchor A-K24), so a carrier signed with a trusted name but a disagreeing provenance falls back to the generic card; peer bodies pass the same authority-envelope neutralization core applies (`<task-notification>` etc. are defused) so a colleague's text can never seed the resume dedup ledger. Lane precedence copies the engine renderer's own order, because the model already read the frame in that order and a client ordering of its own would put a card on screen that disagrees with the frame the model saw. Rendering and parsing of the transcript line live in the same module and are round-tripped in both directions, including a body carrying a forged closing tag (a parser fooled there hands half a message to the next row) and a quote inside the sender label (which must not forge a second attribute); the notice lane is deliberately kept **out** of the parser, since recognising it would mean anchoring the `[Cross-session …]` prefix and a user typing that same line would be rendered as engine speech. Hostile carriers are read as own **data** descriptors only and accessors are never invoked at all — `catch` catches throwing, not never returning — proven by a counting getter that must stay at zero calls, alongside a revoked proxy and a prototype-only carrier; and four legacy payload shapes assert the no-carrier path is byte-identical to before, which is the executable form of "zero difference for an older host" |
+| `scripts/run-peer-frame-projection-test.mjs` | The three engine-injected lanes design/385 puts on the **one** `task_notification` carrier, which are not the same kind of thing at all: a delegated child's uplink (`agentMessage`), another session's message drained from this session's own box (`crossSessionMessage`), and a receipt about one of *this* session's own outbound messages (`crossSessionNotice`). The engine renders none of them inside a `<task-notification>` shell, so a client that projects them as the generic completion card shows "background task finished" while the model read a colleague's sentence — two faces describing different events. The discriminator is pinned to the **typed carrier being present**, never to the `summary` text: those carriers can only be minted by the engine's injection legs (the external `notify()` input is a strict subset of the payload and can wear none of them), while `summary` is filled by every notification there is — so anchoring on text would let any background task impersonate a colleague's message by writing `<agent-message from="…">` into its own summary, and a positive control asserts exactly that payload still projects as the generic card. Fail-closed has two tiers rather than one: a broken **required** field (empty `from`, a non-string `body`, a notice `kind` outside the closed set) returns absence so the caller falls back to the generic card — an honest downgrade where the user still sees the notification — while a broken **optional** field drops only itself, because losing an attribution note and losing a colleague's whole message are not the same magnitude. The provenance side record is **required and must agree on four points** (`kind` matches the lane; `from`/`taskId`/`seq` are present and equal the carrier/payload — each equality is anchored on a core mint site and pinned by the cli wire-anchor A-K24), so a carrier signed with a trusted name but a disagreeing provenance falls back to the generic card; peer bodies pass the same authority-envelope neutralization core applies (`<task-notification>` etc. are defused) so a colleague's text can never seed the resume dedup ledger. Lane precedence copies the engine renderer's own order, because the model already read the frame in that order and a client ordering of its own would put a card on screen that disagrees with the frame the model saw. Rendering and parsing of the transcript line live in the same module and are round-tripped in both directions, including a body carrying a forged closing tag (a parser fooled there hands half a message to the next row) and a quote inside the sender label (which must not forge a second attribute); the notice lane is deliberately kept **out** of the parser, since recognising it would mean anchoring the `[Cross-session …]` prefix and a user typing that same line would be rendered as engine speech. Hostile carriers are read as own **data** descriptors only and accessors are never invoked at all — `catch` catches throwing, not never returning — proven by a counting getter that must stay at zero calls, alongside a revoked proxy and a prototype-only carrier; and four legacy payload shapes assert the no-carrier path is byte-identical to before, which is the executable form of "zero difference for an older host". A re-supplied cross-session message — same task id, status and sequence as the first delivery, handed to the model again after compaction — is not rendered a second time, because the first delivery is still on the user's screen; a record with the next sequence number still renders |
 | `scripts/run-wiring-manifest-projection-test.mjs` | The two end-user facts carried on the engine's `wiring_manifest` frame (`modelGate`: which tools this run's model gate removed and the verbatim restore hint; `autoMode`: whether auto mode is actually armed and the engine's own reason word). Projection: both sections ride as `_sema_`-prefixed superset keys, verbatim, and no SDK-named key is minted; a frame where neither section is well-formed projects to `none/not_in_slice` (no empty arm); `modelGate` needs all three keys and treats `removed: []` as a bad value rather than a reading; `autoMode` needs a boolean plus a non-empty reason that agrees with it, and the reason word is never mapped onto the capabilities vocabulary; the frame is flat (a nested `manifest:{}` wrapper is not a supply); `eventId` rides like every other arm. Adapter: exactly one chrome event on the main lane, a sub-flow frame (any `parentToolCallId`, `null` included) yields nothing, and an absent `eventId` leaves the key absent. Added at receiving time because the shell-side gate could not see this package's behaviour: two mutations (empty `removed` accepted, sub-flow gate removed) had passed the package suite untouched 0.71.0 adds sections F–I: the fourth/fifth/sixth manifest sections (`tools` via the roster reader, `hooks[]` rows dropped one by one when malformed, `lsp` absent unless `mounted` is a boolean), the `tool_roster_delta` arm (narrowed `delta`, `malformed` when `fromDigest`/`roster` cannot be read, host applies it against its own digest), the `context_usage` arm (finite-gated scalars plus `sections[]` rows dropped one by one), and the `WiringManifestMcpEntryView` rename with `MAX_AGENT_SKILLS` gone from the surface |
 | `scripts/run-submit-wiring-manifest-test.mjs` | The non-streaming submit receipt can carry the run's opening wiring manifest (`TaskResult.wiringManifest`, additive on newer servers). `readSubmitWiringManifest` answers one of three: the key is absent on the receipt itself (older server, or a deployment whose engine never produced that frame) — not the same as unreadable; the key is present but cannot be read (not an object, or none of the nine sections survive); or a manifest view. The view is the same shape the streaming lane's chrome event carries (minus its two envelope keys) and is assembled by the same code path, so both lanes agree byte for byte on the same object. Liveness fields ride through untouched — this reader never mints a liveness verdict — and an operator-shaped receipt with extra governance sections reads to the same view as a tenant-shaped one. A zero-tool roster is a real reading, not an absence. |
 | `scripts/run-rule-offers-reader-test.mjs` | The narrowing reader behind the "don't ask again" options, now a public entry point rather than a card-port-only one. Hosts that render the frame themselves (a browser has no three-way terminal card) previously had to rebuild this reader on their side, and what it carries is a **redemption-safety** judgement, not a convenience: the batch arm is redeemed by **index**, so a reader that compacts the array after dropping a malformed entry makes the k-th option a person clicked and the k-th rule the server writes two different rules. So: a bad entry is dropped **on its own** (one bad option must not make a real one disappear) while every surviving entry keeps its **original wire index** — pinned from both ends, with the bad entries leading and trailing. A batch's *members* are the opposite: any malformed member drops the whole batch, because a conjunctive batch is one "yes" to all of them and a batch missing a member is a different grant; its honest-remainder count is a reading, not decoration, so a non-integer or negative value drops the batch rather than rendering a fabricated zero. An empty array, a non-array, an over-cap array and an all-bad array all read as **absence** rather than an empty list, because an empty list renders as "there is an option lane with nothing in it". The two wire generations are ordered by a rule, not a preference: the newer key wins outright, a newer key that is **present but unreadable** does not fall back to the retired key (borrowing the older material would pass someone else's options off as this request's), and a `null` newer key reads as absence so a relaying layer that serialises "missing" as null cannot delete the whole lane on older engines. The public entry is finally reconciled against **both** card-port legs on the same material, byte for byte, so the exported reader and the one the card sees can never become two. Two upstream vocabularies used to be **hand-copied** here, and both had fallen behind: a match word outside the copied pair dropped an otherwise valid option outright, and a batch carrying a directory-read member — a member kind the copy did not know — dropped the whole batch. Both tables now come from one place upstream and are re-exported verbatim, pinned in both directions: every word in the table must be accepted (a narrower copy reds on the words it never learned) and a word constructed to be outside it must still be refused (a reader widened to "any string" reds too), with the retired-key normalising leg sharing the same narrowing so the fix cannot land on one leg only. A member whose kind is genuinely unknown still drops **the whole batch and only that batch** — never one member, because a conjunctive batch one member short renders "yes to N" as "yes to N−1", and never the card, because the honest single beside it is intact — while a member from before the discriminant existed normalises to the historical kind rather than being refused. The additive per-segment reasons ride through verbatim, drop only the row that is malformed, and stay **absent rather than empty** when nothing survives, since an empty list would read as "confirmed nothing uncovered" while the count remains the only source of truth |

package/dist/adapter/activeRunSelfHeal.d.ts

@@ -80,6 +80,7 @@ export type SelfHealOutcome = {
     taskId: string;
     decidePath: string | null;
     decisionInFlight?: true;
+    pendingRowGone?: true;
 } | {
     kind: 'not-parked';
     taskId: string;
@@ -101,6 +102,7 @@ export type SelfHealOutcome = {
     taskId: string;
     decidePath: string | null;
     decisionInFlight?: true;
+    pendingRowGone?: true;
 } | {
     kind: 'ask-decided-without-card';
     taskId: string;
@@ -200,6 +202,8 @@ export declare function attemptActiveRunSelfHeal(signal: ActiveRunBusySignal, ru
 export interface ActiveRunSelfHealCopy {
     wayOut?: string;
     freshSession?: string;
+    keepSessionWayOut?: () => string | null;
+    engineDecidePath?: boolean;
     rowFor?: (outcome: SelfHealOutcome, signal: ActiveRunBusySignal | null | undefined) => string | undefined;
     headlessRowFor?: (signal: ActiveRunBusySignal) => string | undefined;
 }

package/dist/adapter/activeRunSelfHeal.js

@@ -310,6 +310,7 @@ async function planReviewArm(taskId, signal, deps) {
         taskId,
         decidePath: signal.pendingGate?.decidePath ?? null,
         ...(reopenRefusedForDecisionInFlight(verdict) ? { decisionInFlight: true } : {}),
+        ...(verdict.pendingRowGone === true ? { pendingRowGone: true } : {}),
     };
 }
 async function askParkArm(taskId, signal, runs, deps) {
@@ -343,6 +344,7 @@ async function askParkArm(taskId, signal, runs, deps) {
         taskId,
         decidePath: signal.pendingGate?.decidePath ?? null,
         ...(inFlight ? { decisionInFlight: true } : {}),
+        ...(verdict.pendingRowGone === true ? { pendingRowGone: true } : {}),
     };
 }
 async function probeRunStatusOnce(taskId, durable, durableGet, deps) {
@@ -366,6 +368,12 @@ async function staleParkArm(taskId, busy, runs, deps) {
         kind: 'ask-reopen-failed',
         taskId,
         decidePath: busy.pendingGate?.decidePath ?? null,
+        pendingRowGone: true,
+    };
+    const reopenFailedAfterCard = {
+        kind: 'ask-reopen-failed',
+        taskId,
+        decidePath: busy.pendingGate?.decidePath ?? null,
     };
     if (runs === undefined || typeof runs.get !== 'function')
         return reopenFailed;
@@ -406,24 +414,25 @@ async function staleParkArm(taskId, busy, runs, deps) {
         choice = null;
     }
     if (choice === null)
-        return reopenFailed;
+        return reopenFailedAfterCard;
     if (choice === 'wait') {
         rememberDeclinedOffer(declineKey);
         return { kind: 'stale-park-wait', taskId, status: fresh };
     }
     runningChoiceDeclined.delete(declineKey);
     if (callerAborted())
-        return reopenFailed;
+        return reopenFailedAfterCard;
     const afterCard = await probeRunStatusOnce(taskId, runs, durableGet, deps);
     if (afterCard.ghost)
         return { kind: 'ask-run-not-found', taskId };
     if (callerAborted())
-        return reopenFailed;
-    if (afterCard.status === null || !ASK_PARK_STATES.includes(afterCard.status))
-        return reopenFailed;
+        return reopenFailedAfterCard;
+    if (afterCard.status === null || !ASK_PARK_STATES.includes(afterCard.status)) {
+        return afterCard.status !== null && !CLAIM_HELD_STATES.includes(afterCard.status) ? reopenFailed : reopenFailedAfterCard;
+    }
     const stillGone = await readOwnedPendingCount(recheckOwnedPending, deps);
     if (stillGone !== 0)
-        return reopenFailed;
+        return reopenFailedAfterCard;
     const verdict = await cancelAndConfirmRelease(taskId, runs, durableGet, deps);
     if (verdict.outcome === 'released')
         return { kind: 'stale-park-cancelled', taskId };
@@ -605,12 +614,19 @@ export function activeRunSelfHealRow(outcome, signal, copy, origin, follow) {
     const following = follow?.following === true && steerFollowIntent(outcome)?.kind === 'tail';
     if (origin === 'injected')
         return injectedSubmissionRow(outcome, following);
-    const base = activeRunSelfHealBaseRow(outcome, signal, copy?.wayOut ?? DEFAULT_WAY_OUT, following);
+    const engineDecide = copy?.engineDecidePath !== false;
+    const keep = (outcome.kind === 'plan-review-reopen-failed' || outcome.kind === 'ask-reopen-failed') &&
+        outcome.decisionInFlight !== true &&
+        outcome.pendingRowGone !== true
+        ? keepSessionWayOutOf(copy)
+        : null;
+    const base = activeRunSelfHealBaseRow(outcome, signal, copy?.wayOut ?? DEFAULT_WAY_OUT, following, keep, engineDecide);
+    const viaEngine = engineDecide ? decidePathClause(signal) : '';
     switch (outcome.kind) {
         case 'not-parked':
-            return base + lastActivityClause(signal) + decidePathClause(signal) + governanceOriginClause(signal);
+            return base + lastActivityClause(signal) + viaEngine + governanceOriginClause(signal);
         case 'state-unknown':
-            return base + decidePathClause(signal) + governanceOriginClause(signal);
+            return base + viaEngine + governanceOriginClause(signal);
         default:
             return base + governanceOriginClause(signal);
     }
@@ -651,7 +667,19 @@ function decisionInFlightRow(parkedOn) {
         `sema did not show the card again — answering it a second time would send a second decision. It did NOT ` +
         `cancel the run. Your message was NOT sent; send it again in a moment, once the engine has applied that decision.`);
 }
-function activeRunSelfHealBaseRow(outcome, signal, wayOut, following = false) {
+function keepSessionWayOutOf(copy) {
+    if (typeof copy?.keepSessionWayOut !== 'function')
+        return null;
+    try {
+        const v = copy.keepSessionWayOut();
+        const t = typeof v === 'string' ? v.trim().replace(/[.;。]+$/, '').trim() : '';
+        return t !== '' ? t : null;
+    }
+    catch {
+        return null;
+    }
+}
+function activeRunSelfHealBaseRow(outcome, signal, wayOut, following = false, keepSessionWayOut = null, engineDecide = true) {
     switch (outcome.kind) {
         case 'decision-pending':
             return (`This session is held by an earlier turn${outcome.taskId ? ` (run ${outcome.taskId})` : ''} that ` +
@@ -690,22 +718,26 @@ function activeRunSelfHealBaseRow(outcome, signal, wayOut, following = false) {
         case 'ask-reopen-failed': {
             if (outcome.decisionInFlight === true)
                 return decisionInFlightRow(`waiting for your decision (run ${outcome.taskId})`);
-            const viaEngine = outcome.decidePath
+            const viaEngine = engineDecide && outcome.decidePath
                 ? ` You can also decide it on the engine directly: POST ${outcome.decidePath}.`
                 : '';
             return (`The previous turn is parked waiting for your decision (run ${outcome.taskId}) and sema could ` +
                 `not reopen that card here. It did NOT cancel the run — that would have decided it for you. ` +
-                `Your message was NOT sent; ${wayOut} if you no longer want it.${viaEngine}`);
+                (keepSessionWayOut !== null
+                    ? `Your message was NOT sent. To keep this conversation, ${keepSessionWayOut}: the card comes back when the session resumes, if it is still waiting. Or ${wayOut} if you no longer want it.${viaEngine}`
+                    : `Your message was NOT sent; ${wayOut} if you no longer want it.${viaEngine}`));
         }
         case 'plan-review-reopen-failed': {
             if (outcome.decisionInFlight === true)
                 return decisionInFlightRow(`on a plan review (run ${outcome.taskId})`);
-            const viaEngine = outcome.decidePath
+            const viaEngine = engineDecide && outcome.decidePath
                 ? ` You can also decide it on the engine directly: POST ${outcome.decidePath}.`
                 : '';
             return (`The previous turn is parked waiting for a plan review (run ${outcome.taskId}) and sema could not ` +
                 `reopen that approval card. It did NOT cancel the run — that would have discarded the plan for ` +
-                `you. Your message was NOT sent; ${wayOut} if you no longer want that plan.${viaEngine}`);
+                (keepSessionWayOut !== null
+                    ? `you. Your message was NOT sent. To keep this conversation, ${keepSessionWayOut}: the card comes back when the session resumes, if it is still waiting. Or ${wayOut} if you no longer want that plan.${viaEngine}`
+                    : `you. Your message was NOT sent; ${wayOut} if you no longer want that plan.${viaEngine}`));
         }
         case 'running-steered': {
             const handle = `run ${outcome.taskId}`;

package/dist/hitl/toolApprovalWire.d.ts

@@ -128,6 +128,7 @@ export interface ApprovalCardRequest {
     probeCause?: unknown;
     ruleEvidence?: unknown;
     requiresRealApproval?: true;
+    irreversibleParkGate?: true;
     readRootCandidate?: ReadRootCandidateView;
     expiresInMs?: number;
     expiresAtMs?: number;

package/dist/hitl/toolApprovalWire.js

@@ -147,6 +147,8 @@ export async function surfaceFsApprovalAndDecide(deps, taskId, argsByCall, signa
         callKey,
         ...(signal ? { signal } : {}),
         ...(pending.governanceForced === true ? { governanceForced: true } : {}),
+        ...(pending.requiresRealApproval === true ? { requiresRealApproval: true } : {}),
+        ...(pending.gateKind === 'irreversible_ask' ? { irreversibleParkGate: true } : {}),
         ...(pending.hasBidiControls === true ? { hasBidiControls: true } : {}),
         ...(ruleOffersReadOnly !== undefined ? { ruleOffersReadOnly } : {}),
         ...(isWireRecordCarrier(durableProbeCause) ? { probeCause: durableProbeCause } : {}),

package/docs/INTEGRATION-CLIENTS.md

@@ -19,7 +19,7 @@
 
 | 项 | 值 | 真源 |
 |---|---|---|
-| 本包 | `@sema-agent/client-core` **0.82.2**(本批发布版 = **patch**:只增不改的公面 + 若干行为订正 —— 新增措辞单源 `cloudEffectiveNotices`:云控制面有效配置回体的降级告警 / 预算 / 能力上限三读数(CC-164,四态,不读成 0 或「没有预算」);本机引擎 models 文档在宿主注入本机条目判定 `entryAccepted` 时补 `default` / `tierGroups` / `activeTierGroup`,剔掉会让本机整份拒收的目录引用与判定口不收的条目,去掉 registry 注记键 `origin` / `overridesTeam`(CC-165;结果新增 `modelRefIssues`);计划复核结局机读位 `_sema_planReviewOutcome`(CC-160 ①);停泊审批决断操作起止回调 `onDecideOperation` 与操作句柄 `openDecideOperation`(CC-160 ③);新增超集键 `_sema_planReviewOutcome`;peer sdk 地板 `>=11.3.0` 不动;根公面 1229;§94)。上一版 0.82.1 = **patch**:只增不改的公面 + 若干行为订正 —— 新增读口 15 个(续跑折叠码读口与措辞 / 文件历史捕获读口六口 / 模型身份可解析判据五口 / 自定义代理载荷作废重投口 / 决断失败码「门还在」判定)+ 宿主车道声明 `EmitContext.lane`;新增超集键 `_sema_response_model`(终帧 / 每轮用量)与 chrome 用量两臂 `responseModel`、重开判决 `_sema_decisionInFlight`、自愈结局两臂 `decisionInFlight`、内部臂 `_sema_denial_unclassified`、通知码 +1;行为订正:计划复核决断在飞不重开卡且自愈提示不再教人再决一次(CC-160 ②)、决断出口补折叠句且停泊腿不再把折叠 / 坏档误判已决(CC-154)、执行车道 `in-process` 措辞(CC-153)、`-p` 车道 Bash 良性退出与交互同判(CC-168)、`-p` 工具结果帧补结构化结果(CC-149)、被拒分类词两车道同判(CC-172)、非交互车道子代工具帧(CC-158,宿主声明车道才生效)、终帧被拒清单对账流内拒绝(CC-159)、自定义代理载荷作废重投(CC-161);peer sdk 地板 `>=11.3.0` 不动;根公面 1228;§93)。上一版 0.82.0 = **minor**:只增不改 —— 新增只读读口 17 个 + 1 常量、工具名册行一格 `mountedBy`、turn 错误判决 `http` 臂一格 `storedRowWhere`:`/rewind` 代码档与对话档能不能用由引擎能力位回答(CC-146)、云控制面额度读数三分(CC-150)、装配期「这一条腿有没有手」(CC-152)、续跑拒收按原码判(CC-154);两处既有面行为订正:自动模式分类器拒的类别词由 `permission-rule` 改为 `automode-*` 三词(CC-155,订正 0.81.1)、续跑上下文读不出时引擎点名的那一行不再丢(CC-156);0.81.0 预告的 23 键 + 1 子型删改**改到 0.83.0**(§92y ㉙);零 wire 键、零投影臂;peer sdk 地板 `>=11.3.0` 不动;§92。上一版 0.81.1 = **patch**:被拒分类词读器加第③路(CC-148):引擎「没问过就拒」的工具调用(deny 规则 / 权限模式 / hook / 分类器 / 写保护 —— 无 `settlement` 只有 `gate.disposition.deniedBy`)在 user 工具结果记录与终帧 `permission_denials[]` 上此前恒缺 `_sema_denial_kind` / `toolDenialKind`,现按层名落 `permission-rule`(其中「分类器」一项已由 0.82.0 CC-155 订正)(`ask_resolution` 无结算 / 结算段不成形 / 层名表外仍缺席不猜);交互与 `--print` 两条车道同一只读器;另钉终帧两臂 `_sema_effective_*` 键集不变量 —— 成功臂展开单铸函数、错误信封显式重发,覆盖性钉在返回对象自己的类型上(漏键 / 多键编译期即红)+ 门格判两臂键集相等且值逐字节同,给单铸函数加键时错误臂不会再静默少键(本版无键增减);零 wire 键、零投影臂、根公面 1195;§91。上一版 0.81.0 = **minor**:CC 形消息一张脸(C-R85 / CC-131):user 工具结果记录加 SDK 面键 `tool_use_result`(与过渡键 `toolUseResult` 同引用,删驼峰原定 0.82.0、已改到 0.83.0)+ DS-19 普查门升硬门(键集按语法解析、按 type/subtype 分表;23 键 + 1 子型过渡表到期窗原为 0.82.0、已改到 0.83.0,零常驻豁免)+ CC-145 `/rewind` both 臂改发 `restoreFiles:true`、`projectRewind(pending)` 单参;零 wire 键、零投影臂、根公面 1195;§90。上一版 0.80.2 = **patch**:`conflict.resume_context_unavailable` 的 `staleAfterSec` / `runId` 在真 SDK 上读不到(CC-137:sdk 11.1.0 起 `APIError` 把体上非具名键收进 `extra`,这一码映射成 `ConflictError` 无具名位,读口只读顶层 ⇒ 真调用链恒缺席)⇒ 体源结构位改「顶层有效值优先、`extra` 兜底」三只类型化读口单铸,有具名子类位的三处(`retryAfterSec` × 2 / `allowlist`)行为不变只多兜底,头源 `retryAfterMs` 只读具名位;零 wire 键、零投影臂、根公面 1195;§89。上一版 0.80.1 = **patch**:去键重发判据句按「铸」计(CC-143;§87「带键的体恰一次」作废 → 被铸恰一次,POST 几次随传输层)+ 接入文档段落自带射程限定词自 §88 起向前执法 + 模型预设表 deepseek 项改闭集真名 deepseek-flash(CC-144,C-R82 ①)+ 门里 git 兜底自陈探针(CC-142);零 wire 键、零投影臂、根公面 1195;§88。上一版 0.80.0 = **minor**:🔴 peer sdk 地板 >=11.2.1 → >=11.3.0 + 策略拒的归因上 wire(CC-133:发送面 deny+policy 带 settledBy / 读侧 policy_refused 十三词入 CC 规则桶 / 三错误码读口 / 直连门去键重发恰一次)+ 删除规则后果句归包(CC-132)+ /mcp 详情卡两腿分说归包(CC-134:六档读口 + 九句铸点,同批改口八处说满的话)+ 🔴 终帧旧拼法 `structuredOutput` 按 0.79.1 的宣告删除(只剩 CC 形 `structured_output`);根公面 1195;§87。上一版 0.79.1 = **patch**:终帧结构化产出同帧铸 CC 形 `structured_output`(旧拼法 `structuredOutput` 过渡一版,0.80.0 删;CC-130)+ 写审批 `argsOmitted` 退化卡在无受护写入部署形上重新解得出路径(CC-129)+ 云控制面十值二型经本包第二个入口 `@sema-agent/client-core/registry` 原样转口(根入口零变;CC-128 ①);根公面 1184;§86。上一版 0.79.0 = **minor**:server 7.93.0 过境批 —— peer sdk 地板 >=11.2.1 + 写口回执第六臂 unverified 与 503 体 committed 分臂(CC-124)+ 无 run 可读的 MCP 状态面(CC-74)+ 审批卡 mandated 位(CC-110)+ 请求装配第三车道 utility 与 approverPosture / maxTokens / outputRetries 三键 + print 车道 model 座(CC-116 / 118 / 123)+ 拒因两面分铸与四个机读位(CC-125)+ 模型轮边界划消息界(CC-126)+ 委派结果人类面读口按来源声明体内定位(CC-127)+ sdk 转口三件(CC-128)+ 通告码 memory.project_scope_absent(CC-122);公面 1184;§85。上一版 0.78.2 = **patch**:门的棘轮登记数字单真源 —— 五只门十一格棘轮 + 218 条沿革账搬进 scripts/registry.json、门只读不抄、负控改篡改登记物、下调通道 RATCHET-SLACK、新门 run-registry-test(CC-79 第一期)+ 两格存量松量收掉 + submit-wiring-manifest 门补 count===1 边界格;src 零改动,dist 逐字节同 0.78.1;公面 1154;§84。上一版 0.78.1 = **patch**:两辆并行车收货 + 三张跟进票 —— displayBody 人类面读口 + 围栏解析头尾配对收紧(CC-113)/ MCP 两处「没报」措辞分句(CC-115)/ 记忆治理面五动词调用口(CC-97 b)/ 非流式提交回执接线回执读口 readSubmitWiringManifest(CC-121)/ 三份失败判官抛出物读取收口(CC-119);公面 1154;§83。上一版 0.78.0 = **minor**:peer sdk 地板 >=9.8.1 → >=11.0.1(CC-106)+ 单步写口 200 体按 stillLive 判别联合改读、第五臂 revoked、三型取自 sdk、RulesFacade.write 必填、PersistedRuleLiveness 转口(CC-103)+ DeniedBy 第十词 read_boundary 专句(CC-106 ④)+ excludeAllTools 车道行;型面 BREAKING 三处、行为面同 0.77.2;公面 1143;§82。上一版 0.77.2 = **patch**:出包面结构卫生 —— 构建期 removeComments,dist js / d.ts 零注释(新门 run-dist-comments-test.mjs + 卫生门词表 32 条 + 文档向前执法;CC-100)+ gateDeniedByDetail 表外词兜底句改「本版不认识这个层名」(CC-111 ①);行为 / 型面逐字节同 0.77.1;公面 1143;§81。上一版 0.77.1 = **patch**:外部验收三件回填 —— 读器族边界残留 projectMcpReconnectCapability 只认自有键(CC-107)/ rewind 四子键恰读一次(CC-108)/ 子代输出合并按回体 output.type · retrieval_status 定处置(CC-109,server ≥7.92.0 cursorSemantics 缺席面);型面 additive 一处;公面 1143;§80。上一版 0.77.0 = **minor**:请求装配逐键去向回执 + print 车道补座 + 表外键响亮拒(CC-101,四条行为 BREAKING)/ 会话策略写端口(CC-105)/ 持久规则单步写面(CC-103)/ workflow 子代面板行代际三向恢复(CC-104,零改码)/ 能力位读器族口径整族改齐 + 外部验收 F-A / F-B 处置 + 合并树异源复审两轮处置;公面 1143;§79。上一版 0.76.2 = **patch**:制品卫生 —— 清 dist 重 build 去掉两只已撤出源码的编译产物 + dist ⇄ src 对账门;零源码 / 零型面 / 零行为改动;公面 1122;§78。上一版 0.76.1 = **patch**:能力位读器工厂(CC-75)/ MCP 活性观察读口(CC-36)/ `capabilities.peerLane` · `permissionRulesWrite` 四态读口(CC-102)/ 记忆治理面两只读口 + 三只条目面回体读口(CC-97 a)/ 主车道证明逐次新建(CC-99)+ 修 0.76.0 的 F1 / F2(身份四键一次性快照只修了投影口一侧);公面 1122;§77。上一版 0.76.0 = **minor**:六票一批 —— `prompt_assembled` 投影臂 / 审批 feed 三态(**型面 BREAKING**)/ `ApprovalResolution` 单源联合 / 记忆 spec 读口 / 终态词表上游钉 / `tasks_expand` 退役(**型面 BREAKING**),外加身份四键一次性快照(跨全部内部臂);公面 1081;§76。上一版 0.75.1 = **patch**:面板短命子代反序缓行放不出的残洞(外部复验 D1)+ 同代 running 帧三向定形(B4)+ #969 提货 `spawnName` 透传与 `agentType` 改读诚实来源(CC-95);§75。上一版 0.75.0 = **minor**:压缩的非 compacted 结局上 chrome 面 `compaction_outcome`(新内部臂 —— 自建 print 出口的端先表态)+ peer sdk 地板 `>=9.8.1`;型面 additive(ChromeEvent +1 臂)、公面 1065、零 wire 新键;§74。上一版 0.74.5 = **patch**:core 7.24.0 提货 `_sema_freed_tokens?` + 首见即终态的行零事件(幽灵完成行)+ `isNewEngineAgentPanelCycle` / `approvalOutcomeNoteOf` 上公面;型面 additive、公面 1065、peer 不动;§73。上一版 0.74.4 = **patch**:面板键「先发布的键为准」(tick 归不了一的三序不再两行)+ 前代 UUID 迟到 tick 丢弃 + `readRunCancelContext` 下钻收紧;型面 additive、公面 1061、peer 不动;§72。上一版 0.74.3 = **patch**:面板子代身份归一(tick / end 的 UUID 归到 fleet 尾段,宿主零改)+ run 记录 `cancelContext` 读口与成因四词 + workflow 行 `errorCode` + reopen 第二臂两件;型面纯 additive、公面 1061、peer 不动;§71。上一版 0.74.2 = **patch**:`capabilities.deviceExecutor.management` 四态读口(缺席 ≠ false;`deviceExecutor: false` = 车道缺席;管理动词可用性三态归包);公面 1052;§70。上一版 0.74.1 = **patch**:审批卡无决断收口 + 结局回交;§69。上一版 0.74.0 = **minor**:peer sdk 地板 >=9.7.1 + covers 过境 + 面板周期身份 + 离场读口;§68。上一版 0.73.5 = **patch**:workflow 规模告警判定上收进包(`evaluateWorkflowSize` 三态:未知的规模不当作规模正常;分母口径与帽优先级单源;env 读口纯函数);型面纯 additive、公面 1044、peer 不动;§67。上一版 0.73.4 = **patch**:终帧 `permission_denials` 第三键 ⇐ 同一条流 `tool_start` 入参按 `toolCallId` join(三道闸;join 不上留 sema 载体;🔴 更正 0.73.2「自动升级收口」一句)+ 审批 feed push 侧取件失败的欠账退避重试;型面纯 additive、公面 1033、peer 不动;§66。上一版 0.73.3 = **patch**:CC-61 sdk 9.7.1 提货 + `capabilities.approvalsStreamLive` 四态读口(缺席 = 判不出;`livePendingNeedsReconcile` 只在引擎明说会推时才让对账节拍降兜底)+ CC-60 五条模型面文案锚包侧半场(cursorSemantics 机读位 / result 行注记 / bash 首行两变体与良性槽 / 通知 seed 并集 / parked 臂);型面纯 additive、公面 1032、peer 不动;§65。上一版 0.73.2 = **patch**:CC-59 悬挂 ask 全卡 —— `livePending[].frame`(server ≥7.87.0)两道闸窄读进视图,在场即走流内帧腿同一条路(入参可见、改写可转发),缺席保持盲卡;CC-34 收口(引擎半场随 7.87.0 到货,判据交黑盒);型面纯 additive、公面 1026、peer 不动;§64。上一版 0.73.1 = **patch**:CC-57 `decidePlanReview` outcome 尾句按「决断有没有生效」四态选(单铸点 `planReviewOutcomeTail`;此前本地拒发仍让模型汇报已拒绝)+ CC-56 cancel 的 409 按机器码分家(`cancelConflictFromError` 三形 / 新安全码 `cancel_run_not_running`)与提交面 429 用量窗读口 + 人话;型面纯 additive、公面 1026、peer 不动;§63。上一版 0.73.0 = **minor,型面 BREAKING**:fleet 视图八个数值键改可选 —— CC-52 `FleetTaskView.elapsedMs` / `tokens` / `queuedCount` 与 `FleetWorkflowView.doneCount` / `totalCount` / `failedCount` / `tokens` / `elapsedMs` 不再 `?? 0`(有键 = 真值含真 0,缺键 = 不知道;逐键缺席语义见 §62b)+ 台账补偿链无基数不铸;公面导出名零增减 1019、零 wire 变化、peer 不动;§62。上一版 0.72.14 = **patch**:悬挂的流内 ask 拉半场 —— CC-55 `startApprovalsFeed` 快照加 `livePending` 段并进变化摘要(键缺席 = 未报,不铸空)+ 视图 / 去重账 / 与流内帧腿同链的决断口(入参不可见上卡明说、改写不转发)+ 宿主闸对账节拍 + 待决计数读口(不知道 = null);型面纯 additive、公面 1019、peer 不动;§61。上一版 0.72.13 = **patch**:server 7.86.0 提货 —— CC-46 plan_review `permissionModeAfter` 发口(三选卡双闸:宿主声明只读起步 ∧ 引擎 ≥7.86.0;400 field_conflict 两种处置)+ CC-54 `capabilities.executionLane` 四态读面(键缺席 ≠ 不是本机车道;`toolsRunHere` 判据归包)+ CC-53 detach 版本闸三态;型面 additive(另撤回 0.72.12 的主通道 `absent` 臂,改独立缺席订阅口)、公面 1011、peer 不动;§60;**预告 0.73.0 BREAKING**:fleet 视图数值键改可选。上一版 0.72.12 = **patch**:「缺席被折成正向终态」按类清剿 —— CC-50 M1 头条(fleet 行缺席改发 `absent` 不再铸 `end{isError:false}`)+ 六站一次改齐(缺席 ⇒ `unknown`)+ 结果边界 fail-closed 取舍 + `_sema_outcome` + `Outcome unknown:` 行 + CC-51 M6(default 臂不再谎报 running)+ 普查门 + DS-06;型面 additive、公面 993、peer 不动;行为面三处真变化见 CHANGELOG 取舍段;§59。上一版 0.72.11 = **patch**:core 7.23.0 提货 —— CC-49 通告码 66 → 67 `config.write_protection_unresolved`(operator;兄弟码 `target_view_absent` = 没有视图,这枚 = 有视图解析不出)+ `permission_denials` 自动升级腿通电(`gates[].toolCallId` ⇒ `tool_use_id`;CC-34 引擎半场);型面零变化、公面 989、peer 不动、devDep core ~7.23.0;§58。上一版 0.72.10 = **patch**:CC-41 leader run 终局 `needs_human` + `result.conflict` 三形读面(none 缺席不可分 / unreadable 在场读不懂 / conflict;`salvaged[]` 原样给落盘口;措辞单源不提重试);型面纯 additive、公面 989、peer 不动;§57。上一版 0.72.9 = **patch**:server 7.85.0 提货 —— CC-30 会话内 MCP re-dial 消费口(`reconnectMcpServer` 三态 / unsupported 恒五键 / 交易文案单源)+ CC-39② memory-status `autoConsolidationArmed` 三态 + CC-45 resume 补充句渲保留秒数;型面纯 additive、公面 985、peer 不动;§56。上一版 0.72.8 = **patch**:清账三票 —— `memory:"off"` 请求键(CC-39③,server 7.84.0 S-405)+ `runStream` 第二条中性臂 `Model output error:`(CC-43)+ 注入件撞活 run 自动 steer 零卡(CC-44);型面纯 additive、公面 977、peer 不动;§55。上一版 0.72.7 = **patch**:core 7.22.0 提货 —— 通告码 66 `mcp.server_redialed` + 读器 `readMcpServerRedialed` + `wiring_manifest.readDeny` 第九段 `_sema_readDeny`(缺席 = 没有内置行判这条腿);型面纯 additive、公面 971、peer 不动、devDep core ~7.22.0;§54。上一版 0.72.6 = **patch**:cli L-379 包侧半场 —— `steerFollowIntent` steer 后续听意图三态(tail / after-decision / after-wake)+ `activeRunSelfHealRow` 续听声明第五参;型面纯 additive、公面 969、peer 不动;§53。上一版 0.72.5 = **patch**:server 7.82.1 提货 —— `capabilities.webSearch.backend` 部署默认 WebSearch 后端四态读面(键缺席 = 老引擎判不了 ≠ `none`);型面纯 additive、公面 968、peer 不动;§52。上一版 0.72.4 = **patch**:core 7.21.x 三件经 server 7.82.0 提货 —— `wiring_manifest.autoConsolidation` 第八段 `_sema_autoConsolidation`(缺席 = indeterminate 不折 false);型面纯 additive、公面 962、peer 不动;§51。上一版 0.72.3 = **patch**:core #863 记忆单系统半场 ① —— `projectAgentsForWire` 投影 `agents[].memory`(user|project|local 三词同映射引擎平面 agent:<type>);型面纯 additive、公面 962、零 wire 新键、peer 不动;§50。上一版 0.72.2 = **patch**:workflow 视图两票 —— `LiveWorkflowController.resolution()` 六态(snapshot() 的 null 三义拆开)+ `noteOwnWorkflowRun` own 台账同拍登记;型面纯 additive、公面 961、零 wire 键、peer 不动;§49。上一版 0.72.1 = **patch**:server 7.80.1 提货 —— 409 `conflict.resume_context_unavailable` 结构读数(`staleAfterSec` / `runId`;sdk 9.6.0 尚未带键,经 sdk 今天恒缺席)+ `SkillSpec.baseDir`(只在本机车道带);型面纯 additive、公面 960、peer 不动;§48。上一版 0.72.0 = **minor**:型面 BREAKING —— 39 个零消费导出退出公面(CC-32 退役批,[7419] §45 预告 / cli [7422] 同意 / web-client·desktop 过节拍沉默 = 同意;名单见 §47a S-1);运行期零 BREAKING、公面 956、peer 不动、devDep core ~7.20.1;随批:已退役名账 `removed` + export-liveness 门 G 段 + 冻结账门 ⑧c;§47。上一版 0.71.4 = **patch**:core 7.20.1 提货 —— `ENGINE_NOTICE_CODES` 65 + `wiring_manifest.writeProtection` 第七段 `_sema_writeProtection`;型面纯 additive、公面 995、peer 不动、devDep core `~7.20.1`;§46。上一版 0.71.3 = **patch**:门体系清账批 —— CC-24 export-liveness 门 + 39 个导出退役预告(0.72.0 删,表态制)/ CC-26 `as unknown as` 铸链棘轮 / CC-25 冻结账门 ⑧ 段 / CC-06 B-081 包侧半场(终帧 `_sema_effective_reasoning` / `_sema_effective_memory_scopes` + resume reopen 族 + `routePairingVerdict`);型面纯 additive、公面 986 → 995、peer 不动;§45。上一版 0.71.2 = **patch**:接入方验收两票归包 —— CC-28 B-132 decide 拒体三句去内部车道词 / CC-29 B-133 committedUuids 唯一性(重复 ⇒ 形 f);零型面变化、公面 986、peer 不动;§44。上一版 0.71.1 = **patch**:接入档打回修正 —— CC-27 `ApprovalCardRequest.readRootCandidate?` 帧→卡透传位 + 新导出型 `ReadRootCandidateView`;型面纯 additive、公面 986 不变、peer 不动;§43。上一版 0.71.0 = **minor**:型面 BREAKING 六件 —— peer sdk 地板 >=9.6.0 + agent-types >=0.3.0 / `SkillsListingTruncatedFactsView` 删 `requestedBudgetBytes?`·`laneClamped?` / `WiringManifestMcpEntry` → `WiringManifestMcpEntryView` / `MAX_AGENT_SKILLS` 删 / `decide.workflow_host_not_parked` 退役;运行期零 BREAKING。内容:core 7.20.0 提货(闭集 +5 六十四码)/ 🔴 CC-16 `text_segment_end.committedUuids[]` / CC-08 全五件(tool_disclosure / tool_progress 两 chrome 臂、wiring_manifest tools·hooks·lsp 进切片、tool_roster_delta 接上、context_usage 进切片、`readRootCandidate` 镜像)/ CC-19 `mcpEngineLegPresence` 归包 / CC-20 not_parked clean-cut;公面 986;§42 / 42y 八条)|
+| 本包 | `@sema-agent/client-core` **0.82.4**(本批发布版 = **patch**:只增不改 —— park 行审批卡新增门类位 `irreversibleParkGate`(待决行门类恰为 irreversible_ask,今天即生效,门类事实不是保证);`requiresRealApproval` 在待决行顶层带时透传(就绪,待决列表带出后才生效,今天零行为变化)(CC-180);`ApprovalCardRequest` +1 可选位;peer sdk 地板 `>=11.3.0` 不动;根公面 1229;§96)。上一版 0.82.3 = **patch**:只增不改的公面 —— 自愈提示的「保住对话」出路槽 `keepSessionWayOut` 与交互面引擎决断入口开关 `engineDecidePath`(CC-177;两位都不给时措辞与 0.82.2 逐字节同);自愈结局两臂 +机读位 `pendingRowGone`(重开链证实待决项已不在);peer sdk 地板 `>=11.3.0` 不动;根公面 1229;§95)。上一版 0.82.2 = **patch**:只增不改的公面 + 若干行为订正 —— 新增措辞单源 `cloudEffectiveNotices`:云控制面有效配置回体的降级告警 / 预算 / 能力上限三读数(CC-164,四态,不读成 0 或「没有预算」);本机引擎 models 文档在宿主注入本机条目判定 `entryAccepted` 时补 `default` / `tierGroups` / `activeTierGroup`,剔掉会让本机整份拒收的目录引用与判定口不收的条目,去掉 registry 注记键 `origin` / `overridesTeam`(CC-165;结果新增 `modelRefIssues`);计划复核结局机读位 `_sema_planReviewOutcome`(CC-160 ①);停泊审批决断操作起止回调 `onDecideOperation` 与操作句柄 `openDecideOperation`(CC-160 ③);新增超集键 `_sema_planReviewOutcome`;peer sdk 地板 `>=11.3.0` 不动;根公面 1229;§94)。上一版 0.82.1 = **patch**:只增不改的公面 + 若干行为订正 —— 新增读口 15 个(续跑折叠码读口与措辞 / 文件历史捕获读口六口 / 模型身份可解析判据五口 / 自定义代理载荷作废重投口 / 决断失败码「门还在」判定)+ 宿主车道声明 `EmitContext.lane`;新增超集键 `_sema_response_model`(终帧 / 每轮用量)与 chrome 用量两臂 `responseModel`、重开判决 `_sema_decisionInFlight`、自愈结局两臂 `decisionInFlight`、内部臂 `_sema_denial_unclassified`、通知码 +1;行为订正:计划复核决断在飞不重开卡且自愈提示不再教人再决一次(CC-160 ②)、决断出口补折叠句且停泊腿不再把折叠 / 坏档误判已决(CC-154)、执行车道 `in-process` 措辞(CC-153)、`-p` 车道 Bash 良性退出与交互同判(CC-168)、`-p` 工具结果帧补结构化结果(CC-149)、被拒分类词两车道同判(CC-172)、非交互车道子代工具帧(CC-158,宿主声明车道才生效)、终帧被拒清单对账流内拒绝(CC-159)、自定义代理载荷作废重投(CC-161);peer sdk 地板 `>=11.3.0` 不动;根公面 1228;§93)。上一版 0.82.0 = **minor**:只增不改 —— 新增只读读口 17 个 + 1 常量、工具名册行一格 `mountedBy`、turn 错误判决 `http` 臂一格 `storedRowWhere`:`/rewind` 代码档与对话档能不能用由引擎能力位回答(CC-146)、云控制面额度读数三分(CC-150)、装配期「这一条腿有没有手」(CC-152)、续跑拒收按原码判(CC-154);两处既有面行为订正:自动模式分类器拒的类别词由 `permission-rule` 改为 `automode-*` 三词(CC-155,订正 0.81.1)、续跑上下文读不出时引擎点名的那一行不再丢(CC-156);0.81.0 预告的 23 键 + 1 子型删改**改到 0.83.0**(§92y ㉙);零 wire 键、零投影臂;peer sdk 地板 `>=11.3.0` 不动;§92。上一版 0.81.1 = **patch**:被拒分类词读器加第③路(CC-148):引擎「没问过就拒」的工具调用(deny 规则 / 权限模式 / hook / 分类器 / 写保护 —— 无 `settlement` 只有 `gate.disposition.deniedBy`)在 user 工具结果记录与终帧 `permission_denials[]` 上此前恒缺 `_sema_denial_kind` / `toolDenialKind`,现按层名落 `permission-rule`(其中「分类器」一项已由 0.82.0 CC-155 订正)(`ask_resolution` 无结算 / 结算段不成形 / 层名表外仍缺席不猜);交互与 `--print` 两条车道同一只读器;另钉终帧两臂 `_sema_effective_*` 键集不变量 —— 成功臂展开单铸函数、错误信封显式重发,覆盖性钉在返回对象自己的类型上(漏键 / 多键编译期即红)+ 门格判两臂键集相等且值逐字节同,给单铸函数加键时错误臂不会再静默少键(本版无键增减);零 wire 键、零投影臂、根公面 1195;§91。上一版 0.81.0 = **minor**:CC 形消息一张脸(C-R85 / CC-131):user 工具结果记录加 SDK 面键 `tool_use_result`(与过渡键 `toolUseResult` 同引用,删驼峰原定 0.82.0、已改到 0.83.0)+ DS-19 普查门升硬门(键集按语法解析、按 type/subtype 分表;23 键 + 1 子型过渡表到期窗原为 0.82.0、已改到 0.83.0,零常驻豁免)+ CC-145 `/rewind` both 臂改发 `restoreFiles:true`、`projectRewind(pending)` 单参;零 wire 键、零投影臂、根公面 1195;§90。上一版 0.80.2 = **patch**:`conflict.resume_context_unavailable` 的 `staleAfterSec` / `runId` 在真 SDK 上读不到(CC-137:sdk 11.1.0 起 `APIError` 把体上非具名键收进 `extra`,这一码映射成 `ConflictError` 无具名位,读口只读顶层 ⇒ 真调用链恒缺席)⇒ 体源结构位改「顶层有效值优先、`extra` 兜底」三只类型化读口单铸,有具名子类位的三处(`retryAfterSec` × 2 / `allowlist`)行为不变只多兜底,头源 `retryAfterMs` 只读具名位;零 wire 键、零投影臂、根公面 1195;§89。上一版 0.80.1 = **patch**:去键重发判据句按「铸」计(CC-143;§87「带键的体恰一次」作废 → 被铸恰一次,POST 几次随传输层)+ 接入文档段落自带射程限定词自 §88 起向前执法 + 模型预设表 deepseek 项改闭集真名 deepseek-flash(CC-144,C-R82 ①)+ 门里 git 兜底自陈探针(CC-142);零 wire 键、零投影臂、根公面 1195;§88。上一版 0.80.0 = **minor**:🔴 peer sdk 地板 >=11.2.1 → >=11.3.0 + 策略拒的归因上 wire(CC-133:发送面 deny+policy 带 settledBy / 读侧 policy_refused 十三词入 CC 规则桶 / 三错误码读口 / 直连门去键重发恰一次)+ 删除规则后果句归包(CC-132)+ /mcp 详情卡两腿分说归包(CC-134:六档读口 + 九句铸点,同批改口八处说满的话)+ 🔴 终帧旧拼法 `structuredOutput` 按 0.79.1 的宣告删除(只剩 CC 形 `structured_output`);根公面 1195;§87。上一版 0.79.1 = **patch**:终帧结构化产出同帧铸 CC 形 `structured_output`(旧拼法 `structuredOutput` 过渡一版,0.80.0 删;CC-130)+ 写审批 `argsOmitted` 退化卡在无受护写入部署形上重新解得出路径(CC-129)+ 云控制面十值二型经本包第二个入口 `@sema-agent/client-core/registry` 原样转口(根入口零变;CC-128 ①);根公面 1184;§86。上一版 0.79.0 = **minor**:server 7.93.0 过境批 —— peer sdk 地板 >=11.2.1 + 写口回执第六臂 unverified 与 503 体 committed 分臂(CC-124)+ 无 run 可读的 MCP 状态面(CC-74)+ 审批卡 mandated 位(CC-110)+ 请求装配第三车道 utility 与 approverPosture / maxTokens / outputRetries 三键 + print 车道 model 座(CC-116 / 118 / 123)+ 拒因两面分铸与四个机读位(CC-125)+ 模型轮边界划消息界(CC-126)+ 委派结果人类面读口按来源声明体内定位(CC-127)+ sdk 转口三件(CC-128)+ 通告码 memory.project_scope_absent(CC-122);公面 1184;§85。上一版 0.78.2 = **patch**:门的棘轮登记数字单真源 —— 五只门十一格棘轮 + 218 条沿革账搬进 scripts/registry.json、门只读不抄、负控改篡改登记物、下调通道 RATCHET-SLACK、新门 run-registry-test(CC-79 第一期)+ 两格存量松量收掉 + submit-wiring-manifest 门补 count===1 边界格;src 零改动,dist 逐字节同 0.78.1;公面 1154;§84。上一版 0.78.1 = **patch**:两辆并行车收货 + 三张跟进票 —— displayBody 人类面读口 + 围栏解析头尾配对收紧(CC-113)/ MCP 两处「没报」措辞分句(CC-115)/ 记忆治理面五动词调用口(CC-97 b)/ 非流式提交回执接线回执读口 readSubmitWiringManifest(CC-121)/ 三份失败判官抛出物读取收口(CC-119);公面 1154;§83。上一版 0.78.0 = **minor**:peer sdk 地板 >=9.8.1 → >=11.0.1(CC-106)+ 单步写口 200 体按 stillLive 判别联合改读、第五臂 revoked、三型取自 sdk、RulesFacade.write 必填、PersistedRuleLiveness 转口(CC-103)+ DeniedBy 第十词 read_boundary 专句(CC-106 ④)+ excludeAllTools 车道行;型面 BREAKING 三处、行为面同 0.77.2;公面 1143;§82。上一版 0.77.2 = **patch**:出包面结构卫生 —— 构建期 removeComments,dist js / d.ts 零注释(新门 run-dist-comments-test.mjs + 卫生门词表 32 条 + 文档向前执法;CC-100)+ gateDeniedByDetail 表外词兜底句改「本版不认识这个层名」(CC-111 ①);行为 / 型面逐字节同 0.77.1;公面 1143;§81。上一版 0.77.1 = **patch**:外部验收三件回填 —— 读器族边界残留 projectMcpReconnectCapability 只认自有键(CC-107)/ rewind 四子键恰读一次(CC-108)/ 子代输出合并按回体 output.type · retrieval_status 定处置(CC-109,server ≥7.92.0 cursorSemantics 缺席面);型面 additive 一处;公面 1143;§80。上一版 0.77.0 = **minor**:请求装配逐键去向回执 + print 车道补座 + 表外键响亮拒(CC-101,四条行为 BREAKING)/ 会话策略写端口(CC-105)/ 持久规则单步写面(CC-103)/ workflow 子代面板行代际三向恢复(CC-104,零改码)/ 能力位读器族口径整族改齐 + 外部验收 F-A / F-B 处置 + 合并树异源复审两轮处置;公面 1143;§79。上一版 0.76.2 = **patch**:制品卫生 —— 清 dist 重 build 去掉两只已撤出源码的编译产物 + dist ⇄ src 对账门;零源码 / 零型面 / 零行为改动;公面 1122;§78。上一版 0.76.1 = **patch**:能力位读器工厂(CC-75)/ MCP 活性观察读口(CC-36)/ `capabilities.peerLane` · `permissionRulesWrite` 四态读口(CC-102)/ 记忆治理面两只读口 + 三只条目面回体读口(CC-97 a)/ 主车道证明逐次新建(CC-99)+ 修 0.76.0 的 F1 / F2(身份四键一次性快照只修了投影口一侧);公面 1122;§77。上一版 0.76.0 = **minor**:六票一批 —— `prompt_assembled` 投影臂 / 审批 feed 三态(**型面 BREAKING**)/ `ApprovalResolution` 单源联合 / 记忆 spec 读口 / 终态词表上游钉 / `tasks_expand` 退役(**型面 BREAKING**),外加身份四键一次性快照(跨全部内部臂);公面 1081;§76。上一版 0.75.1 = **patch**:面板短命子代反序缓行放不出的残洞(外部复验 D1)+ 同代 running 帧三向定形(B4)+ #969 提货 `spawnName` 透传与 `agentType` 改读诚实来源(CC-95);§75。上一版 0.75.0 = **minor**:压缩的非 compacted 结局上 chrome 面 `compaction_outcome`(新内部臂 —— 自建 print 出口的端先表态)+ peer sdk 地板 `>=9.8.1`;型面 additive(ChromeEvent +1 臂)、公面 1065、零 wire 新键;§74。上一版 0.74.5 = **patch**:core 7.24.0 提货 `_sema_freed_tokens?` + 首见即终态的行零事件(幽灵完成行)+ `isNewEngineAgentPanelCycle` / `approvalOutcomeNoteOf` 上公面;型面 additive、公面 1065、peer 不动;§73。上一版 0.74.4 = **patch**:面板键「先发布的键为准」(tick 归不了一的三序不再两行)+ 前代 UUID 迟到 tick 丢弃 + `readRunCancelContext` 下钻收紧;型面 additive、公面 1061、peer 不动;§72。上一版 0.74.3 = **patch**:面板子代身份归一(tick / end 的 UUID 归到 fleet 尾段,宿主零改)+ run 记录 `cancelContext` 读口与成因四词 + workflow 行 `errorCode` + reopen 第二臂两件;型面纯 additive、公面 1061、peer 不动;§71。上一版 0.74.2 = **patch**:`capabilities.deviceExecutor.management` 四态读口(缺席 ≠ false;`deviceExecutor: false` = 车道缺席;管理动词可用性三态归包);公面 1052;§70。上一版 0.74.1 = **patch**:审批卡无决断收口 + 结局回交;§69。上一版 0.74.0 = **minor**:peer sdk 地板 >=9.7.1 + covers 过境 + 面板周期身份 + 离场读口;§68。上一版 0.73.5 = **patch**:workflow 规模告警判定上收进包(`evaluateWorkflowSize` 三态:未知的规模不当作规模正常;分母口径与帽优先级单源;env 读口纯函数);型面纯 additive、公面 1044、peer 不动;§67。上一版 0.73.4 = **patch**:终帧 `permission_denials` 第三键 ⇐ 同一条流 `tool_start` 入参按 `toolCallId` join(三道闸;join 不上留 sema 载体;🔴 更正 0.73.2「自动升级收口」一句)+ 审批 feed push 侧取件失败的欠账退避重试;型面纯 additive、公面 1033、peer 不动;§66。上一版 0.73.3 = **patch**:CC-61 sdk 9.7.1 提货 + `capabilities.approvalsStreamLive` 四态读口(缺席 = 判不出;`livePendingNeedsReconcile` 只在引擎明说会推时才让对账节拍降兜底)+ CC-60 五条模型面文案锚包侧半场(cursorSemantics 机读位 / result 行注记 / bash 首行两变体与良性槽 / 通知 seed 并集 / parked 臂);型面纯 additive、公面 1032、peer 不动;§65。上一版 0.73.2 = **patch**:CC-59 悬挂 ask 全卡 —— `livePending[].frame`(server ≥7.87.0)两道闸窄读进视图,在场即走流内帧腿同一条路(入参可见、改写可转发),缺席保持盲卡;CC-34 收口(引擎半场随 7.87.0 到货,判据交黑盒);型面纯 additive、公面 1026、peer 不动;§64。上一版 0.73.1 = **patch**:CC-57 `decidePlanReview` outcome 尾句按「决断有没有生效」四态选(单铸点 `planReviewOutcomeTail`;此前本地拒发仍让模型汇报已拒绝)+ CC-56 cancel 的 409 按机器码分家(`cancelConflictFromError` 三形 / 新安全码 `cancel_run_not_running`)与提交面 429 用量窗读口 + 人话;型面纯 additive、公面 1026、peer 不动;§63。上一版 0.73.0 = **minor,型面 BREAKING**:fleet 视图八个数值键改可选 —— CC-52 `FleetTaskView.elapsedMs` / `tokens` / `queuedCount` 与 `FleetWorkflowView.doneCount` / `totalCount` / `failedCount` / `tokens` / `elapsedMs` 不再 `?? 0`(有键 = 真值含真 0,缺键 = 不知道;逐键缺席语义见 §62b)+ 台账补偿链无基数不铸;公面导出名零增减 1019、零 wire 变化、peer 不动;§62。上一版 0.72.14 = **patch**:悬挂的流内 ask 拉半场 —— CC-55 `startApprovalsFeed` 快照加 `livePending` 段并进变化摘要(键缺席 = 未报,不铸空)+ 视图 / 去重账 / 与流内帧腿同链的决断口(入参不可见上卡明说、改写不转发)+ 宿主闸对账节拍 + 待决计数读口(不知道 = null);型面纯 additive、公面 1019、peer 不动;§61。上一版 0.72.13 = **patch**:server 7.86.0 提货 —— CC-46 plan_review `permissionModeAfter` 发口(三选卡双闸:宿主声明只读起步 ∧ 引擎 ≥7.86.0;400 field_conflict 两种处置)+ CC-54 `capabilities.executionLane` 四态读面(键缺席 ≠ 不是本机车道;`toolsRunHere` 判据归包)+ CC-53 detach 版本闸三态;型面 additive(另撤回 0.72.12 的主通道 `absent` 臂,改独立缺席订阅口)、公面 1011、peer 不动;§60;**预告 0.73.0 BREAKING**:fleet 视图数值键改可选。上一版 0.72.12 = **patch**:「缺席被折成正向终态」按类清剿 —— CC-50 M1 头条(fleet 行缺席改发 `absent` 不再铸 `end{isError:false}`)+ 六站一次改齐(缺席 ⇒ `unknown`)+ 结果边界 fail-closed 取舍 + `_sema_outcome` + `Outcome unknown:` 行 + CC-51 M6(default 臂不再谎报 running)+ 普查门 + DS-06;型面 additive、公面 993、peer 不动;行为面三处真变化见 CHANGELOG 取舍段;§59。上一版 0.72.11 = **patch**:core 7.23.0 提货 —— CC-49 通告码 66 → 67 `config.write_protection_unresolved`(operator;兄弟码 `target_view_absent` = 没有视图,这枚 = 有视图解析不出)+ `permission_denials` 自动升级腿通电(`gates[].toolCallId` ⇒ `tool_use_id`;CC-34 引擎半场);型面零变化、公面 989、peer 不动、devDep core ~7.23.0;§58。上一版 0.72.10 = **patch**:CC-41 leader run 终局 `needs_human` + `result.conflict` 三形读面(none 缺席不可分 / unreadable 在场读不懂 / conflict;`salvaged[]` 原样给落盘口;措辞单源不提重试);型面纯 additive、公面 989、peer 不动;§57。上一版 0.72.9 = **patch**:server 7.85.0 提货 —— CC-30 会话内 MCP re-dial 消费口(`reconnectMcpServer` 三态 / unsupported 恒五键 / 交易文案单源)+ CC-39② memory-status `autoConsolidationArmed` 三态 + CC-45 resume 补充句渲保留秒数;型面纯 additive、公面 985、peer 不动;§56。上一版 0.72.8 = **patch**:清账三票 —— `memory:"off"` 请求键(CC-39③,server 7.84.0 S-405)+ `runStream` 第二条中性臂 `Model output error:`(CC-43)+ 注入件撞活 run 自动 steer 零卡(CC-44);型面纯 additive、公面 977、peer 不动;§55。上一版 0.72.7 = **patch**:core 7.22.0 提货 —— 通告码 66 `mcp.server_redialed` + 读器 `readMcpServerRedialed` + `wiring_manifest.readDeny` 第九段 `_sema_readDeny`(缺席 = 没有内置行判这条腿);型面纯 additive、公面 971、peer 不动、devDep core ~7.22.0;§54。上一版 0.72.6 = **patch**:cli L-379 包侧半场 —— `steerFollowIntent` steer 后续听意图三态(tail / after-decision / after-wake)+ `activeRunSelfHealRow` 续听声明第五参;型面纯 additive、公面 969、peer 不动;§53。上一版 0.72.5 = **patch**:server 7.82.1 提货 —— `capabilities.webSearch.backend` 部署默认 WebSearch 后端四态读面(键缺席 = 老引擎判不了 ≠ `none`);型面纯 additive、公面 968、peer 不动;§52。上一版 0.72.4 = **patch**:core 7.21.x 三件经 server 7.82.0 提货 —— `wiring_manifest.autoConsolidation` 第八段 `_sema_autoConsolidation`(缺席 = indeterminate 不折 false);型面纯 additive、公面 962、peer 不动;§51。上一版 0.72.3 = **patch**:core #863 记忆单系统半场 ① —— `projectAgentsForWire` 投影 `agents[].memory`(user|project|local 三词同映射引擎平面 agent:<type>);型面纯 additive、公面 962、零 wire 新键、peer 不动;§50。上一版 0.72.2 = **patch**:workflow 视图两票 —— `LiveWorkflowController.resolution()` 六态(snapshot() 的 null 三义拆开)+ `noteOwnWorkflowRun` own 台账同拍登记;型面纯 additive、公面 961、零 wire 键、peer 不动;§49。上一版 0.72.1 = **patch**:server 7.80.1 提货 —— 409 `conflict.resume_context_unavailable` 结构读数(`staleAfterSec` / `runId`;sdk 9.6.0 尚未带键,经 sdk 今天恒缺席)+ `SkillSpec.baseDir`(只在本机车道带);型面纯 additive、公面 960、peer 不动;§48。上一版 0.72.0 = **minor**:型面 BREAKING —— 39 个零消费导出退出公面(CC-32 退役批,[7419] §45 预告 / cli [7422] 同意 / web-client·desktop 过节拍沉默 = 同意;名单见 §47a S-1);运行期零 BREAKING、公面 956、peer 不动、devDep core ~7.20.1;随批:已退役名账 `removed` + export-liveness 门 G 段 + 冻结账门 ⑧c;§47。上一版 0.71.4 = **patch**:core 7.20.1 提货 —— `ENGINE_NOTICE_CODES` 65 + `wiring_manifest.writeProtection` 第七段 `_sema_writeProtection`;型面纯 additive、公面 995、peer 不动、devDep core `~7.20.1`;§46。上一版 0.71.3 = **patch**:门体系清账批 —— CC-24 export-liveness 门 + 39 个导出退役预告(0.72.0 删,表态制)/ CC-26 `as unknown as` 铸链棘轮 / CC-25 冻结账门 ⑧ 段 / CC-06 B-081 包侧半场(终帧 `_sema_effective_reasoning` / `_sema_effective_memory_scopes` + resume reopen 族 + `routePairingVerdict`);型面纯 additive、公面 986 → 995、peer 不动;§45。上一版 0.71.2 = **patch**:接入方验收两票归包 —— CC-28 B-132 decide 拒体三句去内部车道词 / CC-29 B-133 committedUuids 唯一性(重复 ⇒ 形 f);零型面变化、公面 986、peer 不动;§44。上一版 0.71.1 = **patch**:接入档打回修正 —— CC-27 `ApprovalCardRequest.readRootCandidate?` 帧→卡透传位 + 新导出型 `ReadRootCandidateView`;型面纯 additive、公面 986 不变、peer 不动;§43。上一版 0.71.0 = **minor**:型面 BREAKING 六件 —— peer sdk 地板 >=9.6.0 + agent-types >=0.3.0 / `SkillsListingTruncatedFactsView` 删 `requestedBudgetBytes?`·`laneClamped?` / `WiringManifestMcpEntry` → `WiringManifestMcpEntryView` / `MAX_AGENT_SKILLS` 删 / `decide.workflow_host_not_parked` 退役;运行期零 BREAKING。内容:core 7.20.0 提货(闭集 +5 六十四码)/ 🔴 CC-16 `text_segment_end.committedUuids[]` / CC-08 全五件(tool_disclosure / tool_progress 两 chrome 臂、wiring_manifest tools·hooks·lsp 进切片、tool_roster_delta 接上、context_usage 进切片、`readRootCandidate` 镜像)/ CC-19 `mcpEngineLegPresence` 归包 / CC-20 not_parked clean-cut;公面 986;§42 / 42y 八条)|
 | peer:wire 契约 | `@sema-agent/sdk` **>=11.2.1**(value-level,非 type-only;**0.79.0 抬版**:11.2.x 声明 `capabilities.mcpProbe`、`mcpCapabilities` / `probeMcp` + `McpProbeFace`、写回执第三支 `stillLive: "unknown"`,11.2.1 再补 `TaskRequest.approverPosture` 与审批帧键 `mandated`(型面 + 运行期键锚),本包按 11.2.1 编译,11.0.1 失去物料见证;**0.78.0 曾抬到 >=11.0.1**:11.x 声明 `DeniedBy` 第十词 `read_boundary`、`rules.write` 200 体按 `stillLive` 判别联合、`RemovalLiveness` / `RuleWriteRequest` / `RuleWriteResult` / `RuleWriteBehavior` 包根导出、`TaskRequest.excludeAllTools`,本包按 11.0.1 编译,9.8.1 失去物料见证;**0.75.0 曾抬到 >=9.8.1**:9.8.x 声明 `Capabilities.deviceExecutor.management?`,9.7.1 失去物料见证;**0.74.0 曾抬到 >=9.7.1**:9.7.x 追平的六处型面是本包 0.73.3 起的编译基线,且 9.6.0 失去物料见证;**0.71.0 曾抬到 >=9.6.0**:`tool_disclosure` / `tool_progress` 两帧型与 `ToolApprovalFrame.readRootCandidate` 自 9.6.0 起;0.69.0 曾抬到 >=9.4.0(`reasoning_end` / `lastLegMcp`),0.60.0 抬到 >=8.4.0;地板由 `scripts/run-sdk-floor-test.mjs` 的 `FLOOR` 与 public-surface 门 C-CC-3「不低于见证到的最低实装版」双钉)| `package.json` `peerDependencies` |
 | peer:会话词汇表 | `@sema-agent/agent-types` **>=0.3.0**(type-only,零运行时;0.71.0 抬版:devDep 与 peer 同步,type-superset 门 ④ 撞名预警对着 0.3.0 的 .d.ts 求值) | 同上 |
 | runtime dep | `diff` ^9.0.0(**唯一**一条;portability 门按**等值**钉死) | `package.json` `dependencies` |
@@ -1491,7 +1491,7 @@ CHANGELOG 0.29.0「已知局限」段与相应 JSDoc 都有成文。**别在读
 | **P-4** | med | **`approval_revoke` 在 SDK union 里连成员都没有**(SDK 顶注:known asymmetry,`Registering it is an open item for the next batch`)⇒ 本包不可能有 case ⇒ 运行期落 `dropped('unknown_arm')`;审批链也看不见它(`isToolApprovalFrame` 只认两帧)。全仓 `grep -rn "revoke"` = **0** | `src/adapter/downstream/eventToSdkMessage.ts` 的 `default` 臂;`src/hitl/toolApprovalWire.ts` 的 `isToolApprovalFrame` | 🔴 **引擎撤卡时本包不会替你撤那张卡** —— 被撤的 ask 会一直留在屏上,直到它自己的 5 分钟 TTL / deny 路径触发。要 revoke 语义的端只能自己接 raw named-SSE 腿并撤自己的卡(`unknown_arm` 的 drop 至少留了一行痕) |
 | **P-5** | 真缺口(未定级) | **durable 重放腿上,一条被重放的 `question` 今天不会再打开覆盖层** —— 覆盖层的入口是 `liveQuestionStore` 的 **live demux 写口**,不是投影函数。交互 REPL 无损,「断线后按 `Last-Event-ID` 续读」场景下是真缺口。补它属**行为面**改动(先要答「重放一条已过 5min TTL 的问题该不该弹窗」),按宪法三问单独走 | `src/adapter/downstream/eventToSdkMessage.ts` 的 `question`/`question_complete`/`elicitation`/`elicitation_complete` 臂(缺口逐字记在该处);`src/liveQuestionStore.ts` 头注(`LIVE-ONLY + SAME-REPLICA … No durable resume anchor`) | `respondToQuestion` 当 best-effort 用(404 = 「已经放掉了」,dismiss,**绝不重试**);重连后的恢复走 **409 自愈树 + 自己的 `/v1/approvals` 列举**,**不要**指望重放的 `question` 帧能弹出覆盖层 |
 | **P-6** | ~~low~~ ✅ 0.75.0 已闭 | ~~`compaction_outcome` 在本切片 `not_in_slice`,压缩失败对用户不可见~~ —— 0.75.0(CC-89)起投为内部臂 + chrome 臂 `compaction_outcome`,转录面仍零行(不编 transcript 形),见 §74 | `src/adapter/downstream/eventToSdkMessage.ts` `compactionOutcomeProjection` / `src/adapt/arms.ts` `compactionOutcomeArm` | 端接 chrome 臂后压缩失败可见;不接 = 同旧 |
-| **P-32** | 在册局限(0.36.0 两 wire 键过境后的**辖域**,非缺陷) @cli @web @desktop | **两键各只有一条腿,别按「两面能对上」写码**。① **`requiresRealApproval`**(#283)与 🆕 **`inputHasBidi`**(#341,0.43.0)**同族**:两位都只在**活卡帧**腿(`ToolApprovalFrame` → `ApprovalCardRequest`);**耐久腿零 stamp** —— sdk 7.1.0 / **server 7.46.0 fixture 重验**的`PendingCheckpoint` / `riskDescriptor` 都没有这两位(server 那份在**卡内**:`risk.requiresRealApproval` 是**恒在布尔**、`inputHasBidi` 是卡上的在场位,都不是行上的键),包侧刻意不猜载体名。🔴 **「帧上没有」与「卡上是 false」是两条不同的陈述,别互相推导**(server `ApprovalCardSchema` 顶注逐字)。② **`checkpointId`**(#285 件2):只在 **409 `conflict.session_active_run` 体 + 其 SSE done 帧**的 `pendingGate`;另外三条 park 读面(`/v1/assistant/inbox`、`GET /v1/approvals`、`/v1/approvals/stream`)**都没有**这一格(后两条要给 checkpoint 表反范式一列 = SQL 面双库门,属另一批) | `src/hitl/toolApprovalWire.ts`(帧腿 stamp / durable 腿刻意零 stamp,两处 JSDoc 逐条写明理由)· `src/adapter/runStream.ts`(`ActiveRunPendingGate.checkpointId` + `pendingGateIsProvablyDifferent`)· 行为钉:`run-durable-card-display-keys-test.mjs` ⑨ 段(四种最像的载体名摆在行上,卡入参一个都不许长出来)· `run-selfheal-reopen-test.mjs` G1/G1b | ① **`requiresRealApproval` 在场 ⇒ 一切自动放行让位**(记住的规则 / `allow_session` / bypass 姿态),**缺席绝不读成 `false`**(缺席 = 不是安全类 ask **或** 老 server,两者同形不猜);durable 卡上今天**恒缺席**,别据此认为「durable 门都不是安全类」。② **`checkpointId` 只做两件事**:落日志/排障关联,以及经 `pendingGateIsProvablyDifferent(prev, next)` 做**单向**判断「id 变了 ⇒ 不是刚才那一行」。🔴 **`false` 是「证不出」不是「同一道门」**,🔴 **绝不**当跨调用去重键(server 选行是**无序 `LIMIT 1`**、同 session 可并存多条 pending 行 ⇒ 同一情形连续两次 409 可能报不同 id;也可能 id 不变而 `activeTaskId` 已换),🔴 **绝不**拿它去 join durable 队列行(必然落空,且落空与 legacy 行同形)。缺席是**三成因合流**(legacy 行 / 值没过 server 信任边界校验 / 整只材料读取失败),读作「不知道这道门叫什么」 |
+| **P-32** | 在册局限(0.36.0 两 wire 键过境后的**辖域**,非缺陷) @cli @web @desktop | **两键各只有一条腿,别按「两面能对上」写码**。① **`requiresRealApproval`**(#283)与 🆕 **`inputHasBidi`**(#341,0.43.0)**同族**:两位都只在**活卡帧**腿(`ToolApprovalFrame` → `ApprovalCardRequest`);**耐久腿零 stamp**(🆕 0.82.4 订正:`requiresRealApproval` 改为待决行顶层带这一位时透传,但今天的待决列表不带,durable 卡上实际仍缺席,见 §96 S-1;另见 §96 S-2 的门类位;`inputHasBidi` 那一半不变)—— sdk 7.1.0 / **server 7.46.0 fixture 重验**的`PendingCheckpoint` / `riskDescriptor` 都没有这两位(server 那份在**卡内**:`risk.requiresRealApproval` 是**恒在布尔**、`inputHasBidi` 是卡上的在场位,都不是行上的键),包侧刻意不猜载体名。🔴 **「帧上没有」与「卡上是 false」是两条不同的陈述,别互相推导**(server `ApprovalCardSchema` 顶注逐字)。② **`checkpointId`**(#285 件2):只在 **409 `conflict.session_active_run` 体 + 其 SSE done 帧**的 `pendingGate`;另外三条 park 读面(`/v1/assistant/inbox`、`GET /v1/approvals`、`/v1/approvals/stream`)**都没有**这一格(后两条要给 checkpoint 表反范式一列 = SQL 面双库门,属另一批) | `src/hitl/toolApprovalWire.ts`(帧腿 stamp / durable 腿刻意零 stamp,两处 JSDoc 逐条写明理由)· `src/adapter/runStream.ts`(`ActiveRunPendingGate.checkpointId` + `pendingGateIsProvablyDifferent`)· 行为钉:`run-durable-card-display-keys-test.mjs` ⑨ 段(0.82.4 起:顶层同名键严格 `true` 带,其余最像的载体名不带)· `run-selfheal-reopen-test.mjs` G1/G1b | ① **`requiresRealApproval` 在场 ⇒ 一切自动放行让位**(记住的规则 / `allow_session` / bypass 姿态),**缺席绝不读成 `false`**(缺席 = 不是安全类 ask **或** 老 server,两者同形不猜);durable 卡上今天**恒缺席**,别据此认为「durable 门都不是安全类」。② **`checkpointId` 只做两件事**:落日志/排障关联,以及经 `pendingGateIsProvablyDifferent(prev, next)` 做**单向**判断「id 变了 ⇒ 不是刚才那一行」。🔴 **`false` 是「证不出」不是「同一道门」**,🔴 **绝不**当跨调用去重键(server 选行是**无序 `LIMIT 1`**、同 session 可并存多条 pending 行 ⇒ 同一情形连续两次 409 可能报不同 id;也可能 id 不变而 `activeTaskId` 已换),🔴 **绝不**拿它去 join durable 队列行(必然落空,且落空与 legacy 行同形)。缺席是**三成因合流**(legacy 行 / 值没过 server 信任边界校验 / 整只材料读取失败),读作「不知道这道门叫什么」 |
 
 ### 7b. 决断链缺口
 
@@ -9599,6 +9599,62 @@ installCoreValuePorts({ resolveAutonomousLoopPrompt, loosenReasons, combinePolic
 
 **包侧缺口:** ① 出包面注释存量(别仓名 / 票号 / 帖号 / 过程词)与卫生门判据太窄(CC-100)—— 方案候拍板(dist 去注释 vs 手清),0.77.2 · ② `DeniedBy` 第十词 `read_boundary` / `teamName` / `stillLive` 三词随 sdk 10 型面在 0.78.0 吸收(CC-106)· ③ `capabilities.mcpProbe` 候 core 7.26.0 #951 → server 7.93.0(CC-74)。
 
+## §96 🆕 0.82.4(2026-09-24)park 行审批卡在待决行带 `requiresRealApproval` 时透传 + 过渡位 `irreversibleParkGate`(CC-180)(patch;根公面零增减;`ApprovalCardRequest` +1 可选位;🔴 S-1 待决列表带出后才生效,S-2 今天即生效)
+
+### 96a. 处置表
+
+| # | 面 | 处置 | 消费方要做什么 |
+|---|---|---|---|
+| S-1 | `surfaceFsApprovalAndDecide` 交给卡口的入参(`ApprovalCardRequest.requiresRealApproval`,park 行那一腿) | **new(就绪,待上游)**(CC-180):待决行顶层带 `requiresRealApproval: true` 时,卡入参带 `requiresRealApproval: true`,与活卡腿同一位。只认严格 `true`;`riskDescriptor` 内、卡内 `risk` 子树里的同名键不读。🔴 **今天的待决列表不带这一位**(上游只在别的读面上带),所以现有部署上 park 行卡的这一位仍恒缺席,本版零行为变化;待决列表带出之后即生效 | 卡口在 `requiresRealApproval` 在场时不给会话级放行(记住的规则 / 本会话允许 / bypass 姿态都让位)—— 活卡腿本来就这样分支的端零改动,待决列表带出之后 park 行自动跟上;在那之前 park 行上读不到这一位,别据此认为 park 行的门都不是安全类 |
+| S-2 | `ApprovalCardRequest.irreversibleParkGate?: true`(park 行那一腿) | **new**(CC-180 过渡位):待决行 `gateKind` 恰为 `'irreversible_ask'` ⇒ `true`;`human` / 缺席 / 别的词 / 非串 ⇒ 缺席。本位是**门类事实**,不是「必须真人过目」的无条件保证:引擎把**首判**即要求真人过目的 ask 在 park 时升成这一门类;安全轴收紧的门也落在这里(静态不可逆的工具、外发类工具,以及被 shell 闸收紧的 Bash —— 开了 shell 闸的部署上,大部分 Bash park 都会带这一位)。已知漏口(上游缺陷,已报):只在存储形变后的**复判**里才提出真人过目要求的 ask,引擎在 park 时丢掉这一要求、停成普通门,两位都不带。**只在 park 行腿铸**,活卡腿不铸(那条腿有真位 `requiresRealApproval`)。不冒充 `requiresRealApproval`:两位并存,各说各的事实。展示 + 放行闸,不改本次决断的任何字节 | 想在 park 行上挡住安全类 ask 会话级放行的端,在卡口对 `irreversibleParkGate` 在场的卡不给会话档(覆盖面宽、方向安全,开了 shell 闸时会收掉大部分 Bash park 的会话档;存储复判那条漏口挡不住);待决列表带出 `requiresRealApproval` 之后优先读那一位,本位照铸 |
+
+### 96b. 缺席语义(逐键)
+
+- (S-1)park 行卡上 `requiresRealApproval` 缺席 = 这不是安全类 ask,**或**待决列表没有带出这一位(今天所有部署都属于后者),**不**读成「已确认不是安全类」。
+- (S-2)`irreversibleParkGate` 缺席 = 待决行门类不是 `irreversible_ask`,**或**这张卡来自活卡腿(那条腿不铸本位),**或**行上没有 `gateKind`(早于门类列的历史行);**不**表示「不需要真人过目」—— 活卡腿读 `requiresRealApproval`。
+
+### 96c. 黑盒判据骨架
+
+- **P-G1**(S-1;时间射程:0.82.3 上 FAIL;🔴 射程:用构造的待决行驱动 —— 真部署上要等待决列表带出这一位才可验):一条工具审批待决行顶层带 `requiresRealApproval: true`,经 park 行出卡 ⇒ 卡口入参 `requiresRealApproval === true`;行上该位为 `false` / `'true'` / `1` / `null` ⇒ 卡入参无这一位;同名键只出现在 `riskDescriptor` 内 ⇒ 卡入参无这一位;行上没有 ⇒ 卡入参无这一位(不编 `false`)。
+
+- **P-G2**(S-2;时间射程:0.82.3 上 FAIL):一条 `gateKind: 'irreversible_ask'` 的工具审批待决行经 park 行出卡 ⇒ 卡入参 `irreversibleParkGate === true`,且 `requiresRealApproval` 不因此出现;`gateKind: 'human'` / 缺席 / 非串 ⇒ 无这一位;活卡帧出卡 ⇒ 无这一位。
+
+### 96y. 🔴 三端升级必读(固定段式)
+
+1. **换钉 `^0.82.4` 之前**:零前置(patch;型面只增一个可选位)。
+2. **逐面**:S-1 卡口按 `requiresRealApproval` 收窄会话级放行的端零改动;待决列表带出这一位之前,park 行上它恒缺席。S-2 要在今天就挡住 park 行上安全类 ask 会话级放行的端,在卡口对 `irreversibleParkGate` 在场的卡不给会话档。强制类判定仍走 `approvalIsMandated(card)`。
+3. **成文改口**:无(§7 P-32 ① 的「durable 卡上今天恒缺席」在现有部署上仍成立,不改口;待决列表带出之后再改)。
+
+**包侧缺口:** 宿主座位型面 `ToolPermissionRequest` 上没有 `requiresRealApproval` / `irreversibleParkGate` 的槽(两条腿都没有,0.36.0 起的既有缺口),经座位接卡的宿主今天看不到这两位。待决列表(`GET /v1/approvals`)今天不带 `requiresRealApproval` —— 要在 park 行上真正收窄会话级放行,需要上游在待决列表上带出这一位;本包这一侧已就绪。`approvalIsMandated(card)` 对 park 卡**今天恒为假**:寄存行上的强制以 `ruleOffersAbsence === "mandated"` 表达(且不覆盖全部工具),但待决列表同样不带 `ruleOffersAbsence`,上游补待决列表时一并带出;待决行上的 `mandated` 孪生位上游尚未铸。在那之前,park 卡上的强制门**没有可读信号**:引擎的强制标记(允许规则清不掉这一问)与真人过目要求是两件事,强制门只有同时落在安全轴或首判即要求真人过目时才会停成 `irreversible_ask`,所以 `irreversibleParkGate` **不是**强制门的兜底。上游缺陷:只在存储形变复判时才提出的真人过目要求在 park 时被丢掉(见 S-2)。上游铸出之后,本包按严格 `true` 透传;`approvalIsMandated` 已认 `mandated === true`,宿主届时无需改读法。
+
+## §95 🆕 0.82.3(2026-09-24)自愈提示「保住对话」的出路与引擎决断入口开关(CC-177)(patch;根公面零增减;`ActiveRunSelfHealCopy` +2 个可选位;`SelfHealOutcome` 两臂 +可选位 `pendingRowGone`)
+
+### 95a. 处置表
+
+| # | 面 | 处置 | 消费方要做什么 |
+|---|---|---|---|
+| S-1 | `activeRunSelfHealRow(outcome, signal, copy)` 的 `copy.keepSessionWayOut?: () => string \| null` · `copy.engineDecidePath?: boolean` · 结局 `plan-review-reopen-failed` / `ask-reopen-failed` 的 `pendingRowGone?: true` | **new**(CC-177):这两句在宿主给出「保住对话的出路」时,写成「Your message was NOT sent. To keep this conversation, <宿主那句>: the card comes back when the session resumes, if it is still waiting. Or <放弃那一路> if you no longer want …」—— 保会话的路在前,放弃的路在后,两条都给;宿主那句是动作半句(小写起头、不带句末标点,句末 `.` / `;` 包会去掉)。以下情形不给、也不调用宿主函数:注入件、决断在飞、宿主用 `rowFor` 整行覆写了该结局、结局带 `pendingRowGone`(重开链已证实这道门在引擎上没有待决项,续开也找不回;plan / ask 重开链证出(只认严格 `true`,决断在飞时两臂同样保留)与陈旧 park 臂**呈卡之前**的保守收口带;清理卡呈上之后的收口不带 —— 用户看卡期间 run 可能转去等另一道门,续开能把那张卡带回来 —— 唯一例外是卡后复读确认 run 已终结);其余结局(含卡没能呈上的 running 形)也不给。返回 `null` / 抛错 / 空串 / 非串 ⇒ 当作走不通,句子不变。`engineDecidePath: false` ⇒ 交互面各行(两句 reopen-failed、not-parked、state-unknown)都不给 `POST <decidePath>` 那半句;取消入口(`POST /v1/runs/<id>/cancel`)不受影响,headless 行不看这一位 | 端侧已在包整行后自拼「退出后续开」一句的,改传 `keepSessionWayOut`,判据(交互会话、非后台、新进程读得回这份会话)留在端里;用户够不着引擎决断入口的端传 `engineDecidePath: false`;两位都不传 ⇒ 零改动 |
+
+### 95b. 缺席语义(逐键)
+
+- (S-1)`keepSessionWayOut` 缺席或返回 `null` = 宿主此刻给不出保住对话的路,**不**表示没有这条路;句子与 0.82.2 逐字节同。
+- (S-1)`engineDecidePath` 缺席 = 给(与 0.82.2 同);只有显式 `false` 才关。headless 面不看这一位。
+- (S-1)结局上 `pendingRowGone` 缺席 = 没证出待决项不在(读不出、复证推翻、清理卡在屏上期间状态可能已变、或本就没问),**不**表示卡一定还在。
+
+### 95c. 黑盒判据骨架
+
+- **P-G1**(S-1;时间射程:0.82.2 上 FAIL):`plan-review-reopen-failed` 结局 + `keepSessionWayOut: () => 'run /exit and then \`sema --resume <id>\`'` ⇒ 那一行含 `To keep this conversation, run /exit and then …: the card comes back when the session resumes, if it is still waiting. Or run /clear …`;`ask-reopen-failed` 同;返回 `null` / 抛错 / 空串 ⇒ 与不传逐字节同;注入件、决断在飞、带 `pendingRowGone` 三形不带,且不调用宿主函数;别的结局不调用宿主函数;不传两位时四种行与 0.82.2 逐字节同。
+- **P-G2**(S-1;时间射程:0.82.2 上 FAIL):`engineDecidePath: false` ⇒ 卡重开不了的两句、not-parked 行、state-unknown 行都不含 `POST <decidePath>`;headless 行不变;取消入口照旧。
+- **P-G3**(S-1;时间射程:0.82.2 上 FAIL):ask 重开链判决 `{ reopened: false, pendingRowGone: true }` 且 run 已 `completed` ⇒ 结局 `ask-reopen-failed` 带 `pendingRowGone: true`,该行即使接了 `keepSessionWayOut` 也不给续开那条路;开枪前复证读回非 0 ⇒ 结局不带 `pendingRowGone`,续开那条路照给;清理卡呈上之后 run 转成计划审阅 park(`needs_review`)或状态读不出 ⇒ 不带;卡后复读确认 run 已 `completed` ⇒ 带;plan 重开链证出同理(只认严格 `true`),决断在飞时 plan / ask 两臂都保留这一位。
+
+### 95y. 🔴 三端升级必读(固定段式)
+
+1. **换钉 `^0.82.3` 之前**:零前置(patch;型面只增可选位;不传逐字节同)。
+2. **逐面**:S-1 在包整行后自拼续开句的端,改传 `keepSessionWayOut` 并删掉自拼那段(自拼那段不删、又传这一槽 ⇒ 屏上两句续开);其余端按自己能不能续开会话决定传不传,够不着引擎决断入口的传 `engineDecidePath: false`;用 `rowFor` 自绘这两句的端读结局的 `pendingRowGone`,带它时不要给续开那条路。
+3. **成文改口**:无。
+
+**包侧缺口:** 「新进程能不能续开这份会话」的判据留在宿主 —— 它依赖转录存储位置、会话 id 形、启动目录这些端侧事实,本包只出措辞与排序。计划复核决断没生效那一句的续开指路仍在端侧,不在本版射程。
+
 ## §94 🆕 0.82.2(2026-09-24)云控制面有效配置回体的降级告警 / 预算 / 能力上限读数与措辞单源(CC-164)+ 本机引擎 models 文档补三键、剔悬空引用与被拒条目、去 registry 注记键(CC-165)+ 计划复核结局机读位(CC-160 ①)+ 停泊审批决断操作起止回调(CC-160 ③)(patch;根公面 +1 函数 `cloudEffectiveNotices` 与 17 个型;`CloudProjection` +4 个输出字段;新增超集键 `_sema_planReviewOutcome`〔计划复核结局队列项〕;`HitlBridge` 构造 +可选第三参)
 
 ### 94a. 处置表

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@sema-agent/client-core",
-  "version": "0.82.2",
+  "version": "0.82.4",
   "description": "Client-side session runtime shared by every sema human client (TUI / web / desktop): sema wire frames (AgentEvent) -> CC session vocabulary (SDKMessage) with dual-plane output (transcript/chrome), deterministic transcript ids, lane discipline as a type, and the notification/dedup ledgers. Every CC-skin shape is collected here so the wire itself stays neutral. Renamed from @sema-agent/wire-cc-adapter (0.1.x).",
   "license": "MIT",
   "type": "module",
@@ -41,7 +41,7 @@
     "@sema-agent/agent-types": "^0.3.0",
     "@sema-agent/core": "~7.29.0",
     "@sema-agent/sdk": "^11.3.0",
-    "@sema-agent/settings-schema": "^5.0.0",
+    "@sema-agent/settings-schema": "^6.0.0",
     "esbuild": "^0.27.4",
     "typescript": "^6.0.2"
   }