mirror of
https://github.com/stablyai/orca.git
synced 2026-09-22 08:02:28 +00:00
cb715898cda7465f4bb236282aa5aaba714a5d78
2222
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
  OrcaWinandm4air
|
4085e1cf60 |
fix(memory): release stale session registries (#21734)
* fix(memory): bound session and lifecycle registries * fix(memory): bound transient filesystem registries * fix(memory): cap path and locale caches * fix(memory): bound runtime recovery registries * fix(memory): bound host mirror gap verdicts * fix(memory): bound shell startup env cache * fix(memory): bound gitlab host context cache * fix(memory): release removed ssh generations * fix(memory): expire cloud refresh replay guards * fix(memory): release retired plugin generations * fix(memory): bound plugin log key retention * fix(memory): bound automation authority generations * fix(memory): bound native chat enrichment cache * fix(memory): bound web session tracking generations * fix(memory): bound codex credential absence paths * fix(memory): bound WSL canonical path cache * fix(memory): bound sparse checkout cache * fix(memory): bound shared directory cache * fix(memory): bound advertised URL scan snapshots * fix(memory): bound automation manager cache * fix(memory): bound web session reorder intents * fix(memory): bound web session focus intents * fix(memory): bound web session handoffs * fix(memory): bound automation dispatch tokens * fix(memory): bound host mirror waiters * fix(memory): bound retained session activity * fix(memory): bound retained session activity * fix(memory): bound web session close intents * fix(memory): bound cloud session cache * fix(memory): bound WSL home cache * fix(memory): bound SSH capability cache * fix(memory): bound trust grant cooldowns * fix(memory): bound WSL auth drain state * fix(memory): bound Linear workspace credential cache * fix(memory): bound local Git capability cache * fix(memory): bound WSL Git environment cache * fix(memory): bound WSL Git environment cache * fix(memory): bound WSL preflight cache * fix(memory): keep hot cache entries warm * fix(memory): preserve generation fences across eviction * fix(memory): close remaining eviction fences * fix(memory): align evicted upstream generations * fix(memory): trim successful capability probes * fix(auth): retain expired refresh replay evidence --------- Co-authored-by: m4air <m4air@Mac.localdomain> |
||
 Jinwoo Hong
|
d5dc7b9cf8 |
feat(mobile): budget the terminal snapshot on serialized bytes and hold live output instead of ending the stream (OTA phase C, C7.3) (#21785)
* fix(mobile): budget the mobile terminal snapshot on the bytes it serializes to (OTA phase C, C7.3, ruling 1) The desktop trims a mobile snapshot to 512 KiB of raw terminal text. A client reading it through the page bridge measures the serialized event against a 640 KiB frame cap, and an ANSI snapshot is mostly ESC bytes, each of which JSON spends six on. Measured here on a colour-dense 80-column screen: the raw budget hands back 465,766 bytes that serialize to 669,268 — 102.1% of the cap — so `deliver` answers `cancel(id, 'overflow')` and the terminal is dead before its first live byte, with no recovery that does not reproduce it. `terminal.subscribe` gains an optional `snapshotByteBudget`. A subscriber that sends one is trimmed against the JSON its payload will really cost: the escaped text, plus the metadata it cannot bound from its own side — a path, the OSC-link list, the pending escape tail. A subscriber that sends none, which is every socket client and every older page, keeps the raw byte rule exactly. No negotiation, and none is needed: the field is additive and optional, so an older desktop ignores it and trims as it always did. The page then still has a snapshot over its cap, the shell still ends the stream with `overflow` (C0.3 stands), and the terminal renders its stream-error state rather than a blank pane. The page derives the number from the cap less the event envelope rather than writing it down, so a cap that moves takes the budget with it. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * feat(mobile): hold and coalesce terminal output instead of ending the stream on the window (OTA phase C, C7.3, ruling 2) The shell's backpressure window ends a stream when the page falls 4 MiB behind. That is right for a stream whose reader can survive a gap and wrong for a terminal, whose reader cannot see the hole a dropped chunk leaves — and the window does not wait for a page to go wrong. Measured by the design: the host produces 70.3 MiB/s of JSON and real xterm applies 2.2 MiB/s, so an ordinary `cat` crosses the window in 62 ms. Replayed here through the real ledger against a page draining at that rate, a 5 MB transcript ends the stream after 85 of 107 chunks plain and after 40 of 107 under `grep --color`. Keyed by method on the shell, since the page cannot pick its own window, `terminal.subscribe` now holds what it cannot send, merges consecutive output in escaped bytes under the frame cap, and delivers as the page acks. Nothing is dropped: merging concatenates, and the only exit that loses bytes is ending the stream, which the page is told about. Both transcripts now arrive whole and in order, in 104 and 81 frames, with the largest frame at 622,551 bytes against the 655,360-byte cap. It ends only on the two things that are not slowness: a page that has acked nothing for 20 s, an order of magnitude above the 1.9 s a full window takes to drain, and a backlog past 32 MiB, which at that drain is about 15 s of catching up. Both reach the page as `overflow`, because the shell is the installed app and its page comes from the desktop, so a reason the page's reader has never heard of is a frame it drops rather than an end it acts on. Which one fired, the coalesced-frame count and the peak pending bytes go to the diagnostic log, which is the device proof's only oracle for any of this. Every other stream keeps the byte window exactly, and an event over the frame cap still ends any stream, terminal or not (C0.3). The landed window cases now name a stream the window still governs, so the two rules are never read off each other. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): narrow the event arm the backlog replay reads A binary event carries no `payload`, so the tests-typecheck ratchet refused the reach into it. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * refactor(mobile): narrow the snapshot serializer to the buffer source it reads The changed-code casting gate refused the test's stub runtime, and it was right to: a service-wide type for a function that calls one method is what made the stub need an assertion. The parameter now says what it needs, and the fixture path is no longer one a machine-path grep reads as a leaked local checkout. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): measure the snapshot budget by building the payload, not by summing fields (OTA phase C, C7.3, ruling 14) Round one summed the escaped text and four metadata fields. The payload a bridged client assembles carries nine more — `kind`, `cols`, `rows`, `requestId`, `displayMode`, `reason`, `seq` and both truncation flags — plus the `type` and `streamId` it adds, the `serialized` key and the object's own braces. So a snapshot this host accepted at exactly the budget, with `truncatedByByteBudget` false because nothing had trimmed it, published over the cap and the stream ended with `overflow` before a byte was painted. Measured here on a screen sized to land exactly on round one's budget: the published payload is 655,446 bytes against a 655,273-byte budget, 173 over, and the frame it makes is over the 640 KiB cap by the same amount. The metadata is now built by one function that `sendSnapshotFrames` and the budget both call, and the budget stringifies the payload that function produces. Nothing is summed and nothing is estimated, so a field added to the frame is paid for by the budget the moment it is sent. Where a value is not yet known — the truncation flags, and `seq` or `requestId` at a site that has not fixed them — it is measured at the widest `JSON.stringify` can write it, which is a bound rather than a guess, and forcing `seq` to a number also opens the three fields it gates so those are counted too. The budget therefore travels with the publication fields, because the payload cannot be built without them. On the page, the event envelope is now derived in one place in the protocol module and read by both the snapshot budget and the shell's own merge budget, so the two cannot drift; the page pins the number it sends and the host's cases name that pin, since the two programs cannot import from each other. The case that re-implemented the host's measure is gone: it could not have seen this, because it was the same arithmetic twice. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): arm a held terminal's silence clock only while something is pending (OTA phase C, C7.3) The invariant is "armed implies waiting on the page", and round one broke it in the one direction that kills: an ack re-armed the clock and the drain that followed emptied the queue without clearing it. A terminal that had delivered every byte and gone quiet — which is what a terminal does between commands — would die on `overflow` twenty seconds later. The clock is now synchronised after every change to the queue, so it is armed exactly while something is held. A rule that only ever arms is a rule that only ever ends more streams. Red-first: with round one's arming, an idle stream whose queue has drained still reports its clock armed, and firing it ends a healthy terminal. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): pin the held-stream cases the rulings name (OTA phase C, C7.3) Six cases nothing covered. Two subscriptions on one shell keep separate backlogs, so a busy terminal cannot end a quiet one. A stream the page unsubscribed mid- backlog posts nothing after, and neither does one that has already ended, however much was still held. A payload that is not output breaks a merge run and keeps its place, because a resize is state the reader applies in order. And the budget boundary is checked on the side that enforces it: a payload at exactly the number the page asks the desktop for is delivered inside the cap, and one the cap cannot hold ends the stream under C0.3. The replay no longer acks unconditionally in its catch-up loop. That was the page behaving better than a page can — it acks on reading frames — and it is what hid the silence clock left armed over an empty queue. The held-stream cases close the window on its frame count rather than on four megabytes of string work. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * refactor(mobile): give the event-envelope derivation its own module (OTA phase C, C7.3) `bridge-envelope.ts` is at its line cap and is the protocol's schemas; what a frame costs around its payload is a derivation over them, and two budgets read it — the snapshot the page asks the desktop for, and the output the shell merges. One module, so they cannot drift and neither file is pushed over its limit. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test: narrow the budget fixtures instead of asserting them The changed-code casting gate refused six `as NonNullable<...>` in the new budget cases, and it was right to: a fixture that serialized nothing is a broken case rather than a null to assert away. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * refactor: give the snapshot payload shape its own module (OTA phase C, C7.3) `terminal-snapshot-publication.ts` crossed the root config's 300-line cap, which mobile's own lint does not apply and CI does. The frame's shape and what it costs a client reading it as one payload is a description the budget and the sender both need, so it is the part that leaves. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix: empty a snapshot the budget cannot fit instead of posting it over (OTA phase C, C7.3, ruling 15) Both trimming loops published the zero-row candidate whatever it measured, and zero scrollback is not a small screen: a wide colour-dense viewport still carries its 24 live rows. A capped subscriber could get one frame over its cap, end the stream on `overflow` and paint nothing — worse than a blank terminal, because a blank one repaints on the next byte of output and a stream that never opened does not reopen. Ruling 15: a budgeted subscriber gets that frame with its text emptied and `truncatedByByteBudget` true, never over and never refused. The raw rule keeps its fallback, so an older page and every socket client are served exactly what they were before. Below the metadata the frame must carry there is nothing left to give up, and that boundary is pinned rather than claimed away. The renderer loop is the same walk reached by a different caller and had no test at all; its runtime parameter is narrowed to the two methods it reads so a case can stub it without a cast. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): report what a held terminal stream did instead of calling it an outlived view The backlog report had no branch in the reporter, so it fell through to the "a view outlived its host" warn and every field it exists to carry was discarded. The key made it worse: keyed by kind alone, one backlog per host was ever logged, and a shell holds one stream per open terminal. That report is the only oracle the coalescing rule has. Nothing crosses to the page saying how much was held or how many frames its bytes arrived inside, and both ways a held stream dies reach the page as `overflow`, because a reason its reader has never heard of is a frame it drops. In production the two rules were indistinguishable. They are now a line each, per stream. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test: give the renderer fixture the source its serializer returns `serializeRendererTerminalBuffer` answers `renderer`, and vitest does not typecheck, so the stub's `headless` passed every run and failed the node typecheck instead. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix: budget the frame the publication actually sends (OTA phase C, C7.3) The budget and the publication were written out twice, five lines apart, and had drifted at every site: a budget for `{kind:'scrollback'}` approved a frame sent as `kind:'resized'` with a `reason` beside it, and the live module budgeted `pending-output-overflow` while sending `renderer-mount-ready`. It held only because the padded `requestId` and `seq` are absent from those frames and more than covered the difference. Each site now builds one object and hands it to both. `displayMode` cannot travel that way and was a third under-measure nobody had named: the subscribe flow re-reads it from the runtime after the snapshot is serialized and before the frame is sent, so no caller can tell the budget which mode the publication will carry. It joins `seq`, `requestId` and the truncation flags as a field taken at its widest. The mode list resolves the constant to `never` if the runtime gains a mode it does not carry, so a new one is weighed here rather than found on a phone. Red-first needed a second attempt: the first fixture had trimming slack, so three extra bytes fit and the probe could not see the defect it was written for. The case now budgets a fixed screen at exactly its `auto` measure, where the margin is the whole of the test. One figure for the overshoot everywhere, with its basis: 169 bytes over the 655,360-byte cap on a frame carrying an 8-character request id, 247 with a 24-character one. Three places said 169 and one said 173. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * refactor(mobile): delete two backlog guards no input can reach Both survived mutation because neither is reachable, and neither became reachable when I tried to write a case for it. `next` narrowed the merge ceiling to one frame, but its only caller, `drainTerminalBacklog`, has already narrowed it: the parameter is what one payload may occupy, not what the window holds, so the second narrowing could never change the answer. The parameter now says so and the class no longer needs the frame size at all. The bound still lives in the caller and is still covered: removing it there reds a delivery case. The merge run also compared stream ids, but a backlog belongs to one subscription and every `data` payload on it carries that subscription's single stream id, so the comparison could not fail. The run still stops at anything that is not output, which is reachable and pinned. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * docs(mobile): record the invariant the deleted stream-id guard rested on The merge run compares no stream ids because it cannot need to: a backlog belongs to one subscription and every `data` payload reaching it carries that subscription's single stream id. Written down where the run is, because the thing that would break it is a change made somewhere else — multiplexing two streams onto one record would merge their output into one payload under the first id. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb |
||
|
Jinwoo Hong
|
f5d2d6e757 |
feat(mobile): carry browser screencast frames over the bridge as base64 (OTA phase C, C6.1) (#21758)
* feat(mobile): carry screencast frames over the bridge as base64 (OTA phase C, C6.1) `bridge-screencast-binary.ts` landed in C0 as the page's half of the binary lane and named C6 as the owner of the encoder that satisfies it. This is that encoder, plus the host honouring `wantsBinary`: a subscribe that asked for binary gets an `onBinaryFrame` on the native stream, and each frame crosses as the envelope's `event.binary` on the same `seq` ledger as the stream's JSON events, because the page acks by that count. The base64 encoder is grouped rather than per byte or per `fromCharCode` window. Its docstring carries the measurement, including the part that contradicts the design note this came from: on V8 the per-byte form is the fastest of the three, not the quadratic one, and the chunked form it was meant to beat is the slowest. The grouped one is here because its cost does not depend on how an engine ropes `+=`, and Hermes is what the shell runs. No new opcode, no `v` bump, no negotiation added: `wantsBinary` is already in the contract and is the negotiation. Over-cap behaviour is unchanged in this commit — a binary event over the frame cap still ends the stream, which is what C6.2 changes. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * feat(mobile): drop an over-cap screencast frame instead of ending the stream (OTA phase C, C6.1) Measured at the pane's own request parameters, a screencast frame exceeds the 640 KiB envelope on a phone layout whenever the page will not compress: JPEG's worst case is 0.545 bytes per pixel at quality 72, so mobile view mode at 780x1424 is 811,289 bytes, 124% of the cap. Ending the stream there blacks out a browser tab for the life of the pane over one frame. So the two kinds of event part at the cap. A JSON event that will not fit still ends the stream with `overflow`, because its reader cannot see the hole it would leave; a screencast frame is dropped and the stream lives, because the next frame is one throttle interval away and the pane is still showing the last one. Both are asserted side by side so neither turns into the other. A drop leaves no other trace: the diagnostic beside it prints once per host, so a stream shedding a frame a second and one that shed a single frame read the same. The host therefore counts them per stream for the diagnostic and keeps a session total, and the shell's dev facts carry that total — the surface that already shows build state, with the line moved into its own module so what it says is pinned rather than inferred from a template. The 12-character build prefix it has always shown is unchanged. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * feat(mobile): name the binary screencast lane as a grant (OTA phase C, C6.1) Ruling 5's negotiation, and the check it asked for first: no reader of a grant is a closed enum, so there is no blocker and nothing an older page has to tolerate. `BridgeGrantsSchema.native` and the shell's manifest reader are both open string arrays, and the shell reader's own docstring already states the degradation — a grant name a build does not know leaves that one route native rather than refusing the bundle. What does constrain the name is the host contract's `GRANT_NAME_PATTERN`: a grant is one camelCase token or a `native.<domain>.<action>` verb with at least two dot segments. So `browser.screencast` and `native.screencast` are both refused, and the lane is `screencastBinary`. `screencast` alone would be wrong: the page can already subscribe to `browser.screencast` and receive its JSON events, and only the binary frames need the encoder. Added to the shell's implemented set, which is the same list `init.grants.native ` offers, so a route declaring it is served by a shell that has the encoder and left native by one that does not. No route declares it here; C7's session route does. The contract-side case is a characterisation pin, not a red-first one: the pattern already admitted this name, and the test records that the two tempting spellings are the ones it refuses. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): check the dropped-frame total through the bridge hook (OTA phase C, C6.1) The hook gained a required `onBinaryFramesDropped` two commits ago and this test kept calling it without one, so the tests-typecheck ratchet went red on that commit — caught here rather than in CI because an exit code was read off a pipeline's last stage instead of the script. Fixed by wiring the callback into the probe rather than by a cast, and with the case that makes the wiring evidence instead of types: a dropped frame raises the total the screen receives, and the stream stays subscribed while it does. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * refactor(mobile): keep the dropped-frame counter with the ledger it belongs to (OTA phase C, C6.1) Declared between a getter and a method, which is not where this class keeps state: the subscription map is at the top and the counter is the same kind of thing. Move only. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): serve the binary screencast lane only to a route granted it (OTA phase C, C6.1) Reported as a gap after C6.1's third commit and ruled on: the host honoured `wantsBinary` from any page, so a route that never declared `screencastBinary` could still make the shell encode base64 on its behalf. That is the hole per-route grants exist to close — the same class as a route granted only `navigate` and `storage` reaching the clipboard. The rule now reads the session's resolved list, which is what its route declared narrowed to what this shell implements, and is the same set `init.grants.native` is built from. So the host offers the lane in `init` exactly when it will serve it. Ungranted is not a refusal. The subscription proceeds and its JSON events cross as before, which is the silence every other grant gives at the call site; a page that reads its own grants never reaches that state. Both branches are pinned beside each other, and `grantsForRoute` is pinned dropping a grant this shell does not implement — granted-but-unimplemented and never-granted arrive at the host as the same absence, so its rule reads one case. The grant name moves into the module that holds the rule reading it, so the two cannot drift. `bridge-host.ts` is at 298 of its 300-line cap after this. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * refactor(mobile): move the page's stream-frame rules out of the host (OTA phase C, C6.1) `bridge-host.ts` reached 298 of its 300-line cap, so the next main merge that touched it would have crossed under CI pressure on someone else's PR. Split deliberately instead, at the boundary the growth came from. `bridge-host.ts` is the host's lifecycle and its dispatch. Opening a stream is the only frame kind whose handling is more than one line of delegation — four refusals and, since C6.1, the binary-lane decision — so it moves whole, and `cancel` and `ack` move with it so all three stream frames are decided in one place. The host's `cancel` arm still chooses between a stream and a request where it always did: a page's `cancel` names one or the other, and splitting that choice would leave half an arm in each module. Counted without blank lines or comments, as the rule counts them: bridge-host.ts 298 -> 270, and the new module is 59. A pure move. No test changed and none was added, which is what makes the existing suites the proof: 45 files and 745 tests green on the same assertions as before. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * feat(mobile): report a page that asked for screencast frames it was not granted (OTA phase C, C6.1) An ungranted `wantsBinary` is not a refusal on the wire, so nothing crosses back: the subscription proceeds and its JSON events cross as they always have. That left a page which did ask getting JSON for the life of the document with no side able to say why. `notify-refused` has covered the equivalent notify case since C0; this is the same shape for the one frame kind that lacked it. The rule now answers a verdict rather than a boolean, because `not-asked` and `ungranted` are the same answer for different reasons and only one is worth reporting. So the decision and the report read one rule, and a page that never asked stays silent — pinned, along with a granted route staying silent, so the line cannot start firing on either. The wire is unchanged and pinned unchanged: the case beside this one still asserts one JSON event delivered and zero error frames. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): reset the dropped-frame total with the host that counts it (OTA phase C, C6.1) Round 1 on #21758, three findings. The real one: the count is per host and the screen's copy was not. A rebuilt host starts its own total at zero, so the screen kept the retired host's number until the new one dropped a frame and then read *lower* — a falling count looks like frames coming back, which is worse than starting over. The hook now announces a fresh count as it builds a host. That also reports zero on the first build, where the screen is already at zero and React bails out of the render; the two hook cases pin that leading zero rather than leave it to be rediscovered. Two docstrings that described nothing: `BUILD_ID_PREFIX_LENGTH`'s stayed behind when the constant moved to the dev-facts module and had drifted above `failureMessage`, and `page-route-policy.test.ts` kept the docstring of the test it replaced above the one that replaced it. Both deleted; the first's text lives on the new module. Red-first for the reset, checked against its final expectations rather than its first: with the one line reverted both hook cases fail on the missing zero, and both pass with it. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): keep the dropped-frame total out of a production build's render path (OTA phase C, C6.1) CodeRabbit's Major on #21758. The total went into React state on every dropped frame in every build, and outside a development build the line that reads it renders null — so an over-cap page re-rendered the whole shell screen up to ten times a second for a fact nobody can see. Measured, not argued: five drops, five extra renders. Fixed at the seam rather than with a ternary at the call site. The dev-facts module owns the line, so it now owns the number behind it and the rule that the number is only state where something renders it. The screen holds no flag and no counter; it asks for both and passes the reporter on. The reporter is stable, so the bridge host is never rebuilt for it. `isDevelopmentBuild` becomes a call rather than a module constant. A build flag never changes at runtime so this costs nothing, and as a constant the branch was unreachable to anything that did not set the global before the module loaded — which is why the production case could not be written at the screen at all. Also fixed, found while writing that case: the screen test's `usePageHostSnapshot` double returned a fresh object on every render, so the host effect's identity changed each time and the bridge host was torn down and rebuilt on every render of the screen, settling every pending request with it. The real hook holds the snapshot in `useState` and is stable. One object for the file now. This was masking the fold under test — the count reset to zero on every render — and every other case in that file was measuring a rebuild storm. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * perf(mobile): price a screencast frame before encoding it (OTA phase C, C6.1) Round 2 on #21758, two lows. The encode is a base64 pass over the whole image and the window decides whether the frame can be posted at all, so deciding after encoding made a page that had stopped acking pay for every frame the shell then threw away — the reviewer's case is ten 300 KB frames against a closed window, 3 MB encoded and nothing sent. The size is knowable without encoding: base64 is ASCII, so JSON escapes none of it and the frame is its header serialized plus exactly the image's encoded length. `encodeBridgeScreencastFrame` is now built from that header rather than beside it, so the shape measured and the shape sent cannot drift, and the window arithmetic is one rule read before the encode and again on the frame that was. Exact, not conservative, so the drop diagnostic still reports the whole frame and the committed byte pin is untouched. Red-first with the real encoder wrapped in a counter: window full, ten frames, ten encodes before and zero after, with the drop count still ten. An over-cap frame likewise goes from one encode to none. A third case holds the other direction — two carryable frames still encode twice — so the fix cannot pass by encoding nothing. Second low: the dev-facts block sat outside the only `beforeEach` and left `routeGrants` and `client` mutated, inert only because it runs last. The shared setup moves to file level where the mutable dependencies actually live, resets both, and a case at the end of the file pins it — deleting the reset fails there and nowhere else, since nothing else runs after a case that mutates them. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb |
||
  NeilandXiro The Dev
|
ee354a35d7 |
feat(agents): add OpenCode 2 beta support (#21418)
* feat(agents): add OpenCode 2 beta support Co-authored-by: Xiro The Dev <lethanhtrung.trungle@gmail.com> * fix(opencode2): support current plugin lifecycle and session storage * fix(opencode2): preserve lifecycle ordering and full session capture * test(opencode2): cover setup event bridge * test(opencode2): cover setup event bridge * test(browser): satisfy anti-slop naming check * test(opencode2): cover live form lifecycle * fix(relay): preserve OMP config directory selection * test(opencode2): avoid assertions in bridge fixture * fix(rebase): retain OMP resume and fresh launch behavior * test: align upstream OMP resume expectations * test(opencode2): verify rejected form closes waiting state --------- Co-authored-by: Xiro The Dev <lethanhtrung.trungle@gmail.com> |
||
|
a445abadd4 |
fix(browser): bound CDP output for stalled clients (#20949)
* fix(browser): bound CDP output for stalled clients * fix(browser): log CDP outbound overflow before terminating the client The outbound queue terminated the automation client silently on overflow, so the client saw a socket close indistinguishable from a crash. Surface the cap that tripped and the backlog held when it did. The queue dropped its backlog before invoking onOverflow, so the counters were already zero at the callback. Snapshot them first and pass them through. --------- Co-authored-by: m4air <m4air@Mac.localdomain> Co-authored-by: Neil <neil@stably.ai> |
||
|
Neil
|
e4c7632db2 |
perf(terminal): skip kitty scans for plain PTY output (#21643)
* perf(terminal): skip kitty scans for plain output * fix(terminal): keep the kitty scan fast path total for absent chunks The new escape-byte fast path dereferences the chunk before the string concatenation that used to coerce a nullish value, so an unchecked caller now throws instead of no-opping. Normalize once at the top. Also type the AgentTerminalPreview connect mock against the real preload signature, which turns the stale bare-string replay fixture that tripped this into a compile error. |
||
  
|
3e7da29767 |
feat(editor): add opt-in collapsed unchanged regions for file diffs (#11955)
* feat(editor): add opt-in collapsed unchanged regions for file diffs The combined "View All Changes" diff already collapses unchanged lines into expandable bands (DiffSectionBody sets Monaco's hideUnchangedRegions), but a single-file diff opened from Source Control renders the whole file. Reviewing one changed line in a long file means scrolling past everything else. Adds a General > Editor setting, default off, that applies the same Monaco option to the single-file diff viewer. Off keeps today's full-file rendering. The option is always emitted rather than omitted when off: Monaco retains the last applied value across an options update, so dropping the key would strand an open diff in collapsed mode after the setting is turned back off. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(settings): register collapse unchanged search entry * fix(editor): keep diff viewer under line limit * fix(editor): satisfy diff viewer line budget --------- Co-authored-by: Dan Cieslak <dcieslak19973@users.noreply.github.com> Co-authored-by: Claude Opus 5 <noreply@anthropic.com> Co-authored-by: Neil <neil@stably.ai> |
||
 Jinjing
|
93d245e358 |
Sort dev builds by timestamp instead of semver (#21720)
Dev build base versions can move backwards when a branch is cut before the latest main build. Their embedded timestamp is the authoritative "newest" signal for the picker. For dedicated release repos, compare publishedAt timestamps before falling back to semver comparison. |
||
|
Neil
|
b6b974cb82 |
fix(terminal): clear stale agent identity after shell exit (#21714)
* fix(terminal): clear stale agent identity after shell exit * test(identity): update resolver decision table |
||
|
Neil
|
b0cbb919ba |
fix(pi): use Pi configured provider for Source Control AI defaults (#21693)
* fix(pi): let Source Control AI use Pi configured default When Orca runs Pi for automatic branch names or commit messages without an explicit model override, omit --model so Pi resolves its configured provider. Preserve explicit discovered model selection and add regression coverage. * fix(pi): preserve discovered fallback for non-Pi agents Keep the configured-default sentinel behavior limited to agents whose default is the explicit CLI sentinel. Other dynamic agents still fall back to the first discovered model when their static default is unavailable. * test(pi): pin configured-default dry-run arguments Prove Source Control AI does not render the Pi configured-default sentinel as a literal model argument, and assert explicit model flag pairing positionally. |
||
|
Neil
|
e8a7be4ce2 |
fix(omp): recover retired pane status with validated restart authority
Merged after fresh run 35448889017 passed all required checks, including static analysis, typecheck, package jobs, all test shards, changed E2E, Docker SSH E2E, and verify. |
||
|
Neil
|
e7da72c3d7 |
fix(omp): attach desktop and mobile images through file mentions
Merge fully verified: desktop/mobile focused suites, node and mobile typechecks, changed-code quality, hosted RPC recording pin, package checks, all test shards, and verify pass. This fixes #20389 across composer, drop, picker, and mobile clipboard-accessory paths. |
||
 Brennan Benson
|
061a756b84 |
test(agent-status): pin that omp's approval_mode cannot hide a real prompt (#21499)
omp forwards its `approval_mode` on every `tool_approval_requested`, and the shared normalizer deliberately ignores it. Nothing recorded why, so the field reads like a dropped qualifier that a future change should start honouring. It must not be honoured. Measured against omp 17.0.5: the CLI emits this event only after its own policy engine already resolved the call to "prompt", and then parks on a human Approve/Deny select. Auto-approved calls emit nothing at all. `approval_mode` carries the ambient mode (always-ask | write | yolo), not the verdict, so a per-tool `tools.approval.<tool>: prompt` produces a genuinely blocked human carrying `yolo` -- the one value that looks auto-approving. No behaviour change. Records the reason at the decision site, replaces two fixtures that asserted an `approval_mode` of 'prompt' (not a member of omp's enum) with captured values, and adds guards pinning that every real mode, plus a missing or unrecognised one, stays blocked, and that pi is unaffected. |
||
 NeilandSudoAI-DEV
|
ae9c06c941 |
feat(omp): discover and switch native-chat models on desktop and mobile (#20612)
* fix(omp): discover and switch native-chat models Report the running OMP provider/model and discover available choices on the execution host for desktop and mobile. Register an extension command to switch through the OMP API because its TUI does not accept /model args. Advertise that command in status so older hosts remain read-only. Addresses the OMP portion of #17603; Pi chat enablement remains separate. Model reporting begins on lifecycle activity; no startup status is invented. Co-authored-by: SudoAI-DEV <220139811+SudoAI-DEV@users.noreply.github.com> * refactor(omp): check generated model metadata types * test(omp): verify model picker command and reported selection * test(omp): add repeatable real model-switch runtime proof * test(omp): require model capability delivery in runtime smoke * fix(mobile): decode OMP model discovery through RPC operations * fix(omp): preserve exact reported model selectors * fix(omp): preserve generated extension syntax after rebase * fix(omp): merge generated harness UI context types * test(omp): model switching keeps one session manager * test(omp): include transcript path in model status proof * test(omp): avoid renderer error-type union * test(omp): keep renderer test type explicit --------- Co-authored-by: SudoAI-DEV <220139811+SudoAI-DEV@users.noreply.github.com> |
||
|
Jinwoo Hong
|
c22c442fdb |
feat(mobile): answer native verbs on the shell, clipboard first (OTA phase C, C2.4) (#21623)
* feat(mobile): declare the native verb table and advertise it (OTA phase C, C2.4) The contract half of the shell-answered request seam: the `native.` prefix, a typed table with params and result schemas per verb, and the two clipboard verbs. `MOBILE_WEB_SHELL_GRANTS` spreads the table's own name tuple rather than restating it, so a verb cannot be advertised without a row and a row cannot exist unadvertised — the table is `Record<BridgeNativeVerb, …>`, so a missing row does not compile, and the suite holds the other direction. Verb names go in the flat grant list on purpose: a route may declare one, and a shell that lacks it keeps that route native rather than walling it. The mime shape admits `image` because a later build will serve one; this one refuses it, and the reason will say out of scope rather than unsupported, since `expo-clipboard` implements the image calls. No frame kind is added and no protocol version moves. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * feat(mobile): answer native verbs on the shell and fence them from the desktop (OTA phase C, C2.4) The host half of the seam. `forward()` is the one place a request reaches the client, so the `native.` check sits there and nothing about ids, caps, settlement or cancel moves: a native request takes a pending slot and settles on the same frames as a forwarded one. `readBridgeNativeVerbCall` is the whole decision, separate from the host so the `ungranted` arm can be exercised at all — every page is offered every verb this build implements, so through a real host that arm is unreachable today and is the point of the check once a grant is per-route. Refusals carry `native_verb_refused`, which the desktop's vocabulary does not contain: an unlisted method comes back from `MOBILE_RPC_METHOD_ALLOWLIST` as `forbidden`, so reusing that would make a leaked fence read as an ordinary scope refusal. Every case in the host suite reads `client.requests` for the same reason. `_meta` is omitted from host-authored replies per the ruling, which required making it optional on `RpcSuccess`/`RpcFailure`: the type required a field the wire never has. `isRpcResponse` does not read it, `runtime-rpc-envelope` already makes it optional on a failure, and nothing in this app reads it — every occurrence is a fixture writing one. Zero other type errors resulted. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): restore the harness verb-type import and drop an unused one Two leftovers from threading the native reply type through and then removing it: the host harness lost its `BridgeNativeVerb` import, and the request module kept a type import nothing uses. `tsc` and oxlint both failed on the previous commit; this is the follow-up rather than an amend. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): typecheck the native verb suites and drop the dead reply type Three leftovers the ratchet caught, none visible to `tsc -p tsconfig.json`, which excludes test files: - the fence suite read `frame.payload` off the whole `reply` union, and a chunked reply has no `payload`; it narrows on the field now - the bridge hook's own suite builds its caller options inline and had no `serveNativeVerb` - `BridgeHostAuthoredReply` became unused once `_meta` was optional, and an exported type nothing reads is the pattern round 2 of C2.3 flagged; the statement it carried already lives in the verb table's header Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * feat(mobile): give the page a typed surface for the native verbs (OTA phase C, C2.4) `useNativeVerbs` is the page's side, typed from the same table the host serves, so a verb cannot be called with params the shell will refuse. Each call goes out as an ordinary `request` and settles on the ordinary frames; the method name is the whole difference. A verb the shell did not grant is refused before a frame is sent, because a rejection after a round trip and one that never left look identical to an `await` and only the first costs an in-flight slot — `granted` is exposed so a caller can pick its own fallback instead. Results are parsed rather than trusted: the shell is a different build than the page, and a result shape that moved should fail at the seam rather than halfway through a screen reading a field that is not there. No call site uses it yet; the two `Clipboard.setStringAsync` sites are the consumer PR's. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): send native verbs from the module that owns the request port (OTA phase C, C2.4) `use-native-verbs.ts` called `client.sendRequest` directly, which the unvalidated-request-port boundary refuses: new code must send through an `RpcOperation`, and nothing may be added to the inventory. An `RpcOperation` is not available to this seam. Its `method` is typed `RpcMethodName`, which is `keyof typeof RPC_PARAMS_BY_METHOD` from the desktop's generated params catalog. Putting `native.clipboard.read` there would declare that the desktop serves a method the whole fence exists to keep off it. So the send moves into `bridge-rpc-client.ts`, already listed as an owner of the port — a module that implements the port rather than a call site picking its own method and acceptance. `callNativeVerb` rides the same frame, id space and in-flight cap as any request, and the page surface stays a thin typed wrapper that reaches no raw port. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): fence native methods on subscribe, not only on request (OTA phase C, C2.4) The fence sat in `forward()`, which is the one place a *request* reaches the client. A `subscribe` reaches the same client by another door: a frame naming `native.clipboard.read` opened a real stream on the desktop, and because `client.requests` stayed empty the whole suite read as green over it. Refused in `handleSubscribe` before the id is claimed, under the same `native_verb_refused` code, so nothing about the frame reaches the desktop or occupies a slot. Cancel needs no arm of its own: it can only settle an id that was admitted, and none is. The oracle is widened with it. Every case now reads the client's streams as well as its requests, because the old one could not see this at all — an absence that only ever looked at half the boundary. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): hold a native verb's answer to the result it declares (OTA phase C, C2.4) The table names a result schema per verb and the host never applied it, so a handler could answer `{ nonsense: 1 }` and the page's own parse would be the first to notice — halfway through a screen, not at the seam. Validated on the host and refused by name on a mismatch, which is what makes the table's claim true on the side that serves it. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): keep the native verb member from being a raw port (OTA phase C, C2.4) `callNativeVerb(verb: string, params: unknown)` took any method, so `callNativeVerb('worktree.list', …)` reached the desktop through the real pair — a raw request port in the one module allowed to hold one, and invisible to the inventory, whose scan counts `.sendRequest` shapes and not a bare call inside the owner. The parameter is typed `BridgeNativeVerb` now, which is the fence for every caller the compiler can see, and the prefix is checked at runtime for one that reached the member through a widened type. The compile-time half is pinned by a `@ts-expect-error` the tests-typecheck ratchet holds: widening the parameter back makes that directive unused and fails there. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): give every native verb refusal one typed error at the page (OTA phase C, C2.4) Only the seam's own refusal carried `native_verb_refused`. A handler that declined and a reply too large to send arrived as other categories with no code at all, and the hook rethrew a bare `Error(message)` — so a caller telling an out-of-scope mime from an unsendable clipboard had to read message text, and those want different handling. Three changes, one shape. The host re-raises a handler's failure under the seam's code, keeping the handler's message because that is what says why. `BridgeReplyUndeliverableError` carries its frame refusal as a code, so `reply-too-large` survives to the page. The hook throws `NativeVerbError` with a `reason` read off the code `reconstructBridgeError` already copies onto the rejection, plus `ungranted` for the arm this side decides. Removes the unreachable `ok: false` branch from the hook with it. The narrowing it was doing moves into the client member, which now promises a success or a rejection and nothing else. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): test the in-flight cap and cancel, not the duplicate-id branch (OTA phase C, C2.4) The case named for the cap sent the same id twice, so what it exercised was the already-in-flight check. It never held a second slot and would have passed against a seam that took none. It now fills the cap with distinct ids against a handler that never settles, and asserts the one over it is refused with the cap's own message. A cancel case goes with it: a native request cancelled before its handler settles posts nothing afterwards, the way a forwarded one does not answer an exchange the page has moved on from. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): pin the native verb member's type with a directive, not a cast The case proving a desktop method cannot go through `callNativeVerb` reached the runtime guard with `as never`, which the casting gate refuses — and a cast is the wrong tool anyway: it asserts past the very type the case exists to pin. `@ts-expect-error` instead, which the tests-typecheck ratchet holds: widening the parameter back to `string` makes the directive unused and fails there. The call still runs, so the runtime guard is exercised too. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): parse the shell's error code instead of reaching for it (OTA phase C, C2.4) The anti-slop audit refuses `Reflect.get`: dynamic input is parsed into a named shape before it is read. `code` is not a property of `Error` — it is whatever `reconstructBridgeError` copied onto the rejection from the capture — so a schema is the honest reader here, and it says what this takes without asserting the rest away. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): name the id collision before the native fence on subscribe (OTA phase C, C2.4) The fence ran before the already-in-flight check, so a `subscribe` naming a `native.` method under a live request's id settled that request with the fence's message. The page lost the request either way — the collision class predates this PR — but it was told the wrong cause, which is the difference between a bug it can see and one it cannot. Collision first. The fence still runs before any slot is taken, so nothing about the frame reaches the desktop. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * refactor(mobile): parse a native verb result once, inside the catch (OTA phase C, C2.4) The result was parsed twice: by the send path against the table's schema, and again at each caller against the concrete one. The second parse was dead, and it sat outside `call`'s catch, so a shell answering a shape the page did not expect would have escaped as a bare `ZodError` — the one shape this surface promises not to throw. `call` takes the verb's result schema and parses once, inside the catch, so every failure leaves as a `NativeVerbError`. The params parse at the callers goes with it; the host validates params and the page builds them typed. Also moves the comment block documenting `onExternalLink` back above it, which `serveNativeVerb` had landed in front of. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): give every native verb refusal its own code, and keep handler words on the device (OTA phase C, C2.4) Two findings that had to land together. Six faults all arrived as `native_verb_refused` and differed only in message text, which the hook's own comment said nobody may switch on. And a handler's message crossed verbatim: a clipboard read that failed after reading is free to put what it read in its error, and the error frame is the only path out of this seam that is not a declared result. So each fault gets a code — unknown verb, ungranted, bad params, wrong result, out of scope, handler failure, native-on-subscribe — and the three paths that reached the page uncoded get one too: the in-flight cap, a non-native method through a widened member, and host disposal. `reason` is now drawn from a declared list with no `unknown` arm, asserted at the hook. A handler's code crosses and its message does not; the shell logs the real one. The out-of-scope mime stays distinguishable because the code carries it, not the text. `bridge-host.ts` crossed the line cap with this, so the serving half moves to `bridge-host-native-verbs.ts` — read the call, serve it, hold the answer to what the verb declares — leaving the host the frames around it. No cap was disabled or raised. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): refuse unknown verb params, and floor an unknown code (OTA phase C, C2.4) Two the bots caught, both about a shape one side does not know. `z.object` strips unknown keys, so `{ mime, value, unexpected }` dispatched as if the extra key had not been sent — and the page and the shell are separate builds, so a param the shell silently ignores is the shape of a verb that changed underneath a page. `z.strictObject` on the verb params and results. And the page passed any code through as `reason`, while its own doc and `NATIVE_VERB_REASONS` promised a closed list; a shell newer than the page would have fallen off the end of a caller's switch. Unrecognised codes floor to `unreported`, `reason` is typed to the list, and the doc says what the list actually is rather than the single code the per-arm ones replaced. The flooring is tested by delivering the frame such a shell would send: this build's host normalises an unknown code before it leaves, so the pair cannot produce one. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): serve no request before the page has asked for a session (OTA phase C, C2.4) `serving` starts true so a page's first frames are not refused for arriving in the same native batch as its `ready`, but nothing checked whether an `init` had ever been sent. So a request from a document this host had told no caps, no grants and no route was forwarded to the desktop, or served as a native verb, while the notify path had refused exactly that since C0. Gated on `initSent`, under the protocol's own `before-ready` name. Streams are left alone: the finding names requests, and gating `subscribe` too is a wider change than it asked for — worth its own decision, since the same hole is there. Fourteen host cases and five hook cases were relying on this: they open a request without ever asking for a session, which no real page does. They take a `ready` now, through a harness option, and the counts that read what the host posted account for the `init` a session opens with. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): grant a page session what its route declared, not what the app can do (OTA phase C, C2.4) `init.grants.native` handed every session the shell's whole capability set, so a route declaring only `navigate` and `storage` was granted `native.clipboard.read` as well. That was harmless while every grant was a navigation or a write the page could make anyway. It stopped being harmless the moment a verb reads something back, which is this PR. The session is now granted the intersection of what this shell implements and what the mounted route declared in `MOBILE_WEB_PAGE_ROUTES`, plus the protocol's own `fault`. One list: `init` issues it and every grant check — notify and native verb — reads the same one, so what a page is told it may do and what it will be served cannot drift. `MOBILE_WEB_SHELL_GRANTS` and `implementsGrant` are unchanged; the shell's capability set is still the ceiling a route's list is drawn from. User-mediated authorization is not attempted here and goes to C2.7 as an open question. `use-mobile-web-shell-session.ts` crossed the line cap with the extra field, so the three effect workers that touch the network and the disk move to `mobile-web-shell-session-effects.ts`, leaving the hook its reducer and callbacks. No cap was disabled or raised. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): key the pre-handshake refusal on the session, and apply it to streams (OTA phase C, C2.4) Two rulings, one mechanism. The gate keyed on the host instance, and a host is rebuilt whenever the client under it changes. The page does not know: the session id is the same, so it neither re-handshakes nor hears that the shell was replaced. So a live page's next request was refused, and would have been until reload — a regression, not a safety gain, and not covered by the in-flight settling as delivery-unknown. The host now inherits whether its session already handshook, which the hook records when the page first asks. And the rule is about the session rather than the frame kind, so `subscribe` is gated with `request`: a stream opened before the handshake was the same hole. Fourteen stream cases were exercising a state the protocol forbids — they subscribe without ever asking for a session, which no page does. Every one is about caps, backpressure windows, acks, cancel, idempotency or arity; none was testing anything through the hole itself. They complete the handshake now, and the counts that read what the host posted account for the `init`. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): judge a cached fallback by its own routes and grants (OTA phase C, C2.4) The newer manifest is read before the download is attempted, so its `pageRoutes` and `routeGrants` are already on the session when the download fails. Opening the cached generation then mounted an older page under a newer bundle's grants: a cached route that never declared the clipboard would have been granted it by a manifest it is not running. The fallback now derives both from `cached.routes`, and applies that generation's own render eligibility before mounting it — a route only the newer bundle claims is not a route the cached page can serve. This is the Phase D "grants across generations" item arriving early. Only the grant side is fixed here; persisting a generation's grants with the generation itself stays Phase D's. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): remount the shell on a route change, so its bridge cannot outlive it (OTA phase C, C2.4) A host captures the grants its session was opened with, and the agent-history route renders `MobileWebShellScreen` with a pathname derived from `worktreeId` and no key. So changing worktree updated the screen in place: the old bridge stayed mounted and kept authorising frames under the grants of the route the page had already left. Keyed on the route now, which makes the change a remount — the old bridge is disposed in the commit, before it can read another frame, and the new session starts with no grants until its own `init`. The worktree-list and embedded- browser routes are keyed on the host id for the same reason; the hazard is the same whenever a dynamic segment moves under a mounted shell. The probe that catches this uses an empty dependency array on purpose: keyed on the pathname it re-fires on a prop update and reads exactly like a remount, which is the one thing it exists to tell apart. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(shared): let a manifest declare a native verb as a route grant (OTA phase C, C2.4) `GRANT_NAME_PATTERN` was dotless, and the contract's own pin asserted a dotted grant is refused. So no manifest the desktop can produce could declare `native.clipboard.write` — and once grants are scoped per route, a verb no route can name is a verb no route is ever granted. Every native verb was unreachable for every route. The grammar now admits the verb shape the table names: `native.` followed by at least two lowercase dotted segments, which is `native.<domain>.<action>`. A plain name wearing a dot is still refused, `native.navigate` included, so the pin keeps its meaning. Wire compatibility, checked rather than assumed: widening what a manifest field may contain is a new optional value reaching readers that shipped before it, and the phone's reader already tolerates one. Its route schema bounds a grant's length and nothing else, deliberately — an unknown name is not a parse failure that would refuse the whole bundle, it is a grant this build does not implement, so `implementsGrant` drops it and the route stays native. Both halves are now tested: an unknown verb leaves its route native and grants nothing, and a known one reaches `init.grants.native`. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): split the session reducer suite by concern before main pushes it over the cap Merged with main the reducer suite reaches 805 lines against a cap of 800 — neither side alone crosses it, which is the case the lane rules warn about. Split at a concern boundary rather than raised: the grant-facing cases (the cached fallback's own routes, and a manifest verb reaching the session grants) move to `mobile-web-shell-session-grants.test.ts`, and the fixtures both suites drive the reducer with move to a shared module beside them, the way the bridge host suites already share a harness. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb |
||
|
Neil
|
ff934256ae |
fix(omp): retain recorded transcript paths when resuming (#20634)
Based on the resume-locator proposal in stablyai/orca#16276 by @CodeHourra. Retains UUID-based ownership and existing reattach behavior. |
||
 Neilandunknown
|
4dec64d533 |
feat(source-control-ai): support OMP generation (#20624)
* feat(source-control-ai): support OMP text generation Read prompts on stdin, retain OMP configured model by default, and reuse JSON model discovery. Co-authored-by: unknown <1784931579@qq.com> * test(source-control-ai): cover OMP large input and model overrides * fix(omp): keep configured model default out of discovered catalog * fix(omp): hide config default from model discovery catalog * fix(omp): separate terminal discovery from generation defaults * test(omp): keep model probe import compatible with CLI typecheck * test: align Source Control AI registry contracts with OMP --------- Co-authored-by: unknown <1784931579@qq.com> |
||
|
Neil
|
605a4ef868 |
fix(omp): start new tasks without auto-resuming old sessions (#20622)
* wip(omp): prove fresh settings overlay without redirecting storage * fix(omp): guard fresh launches with execution-host settings * fix(omp): preserve unmodelled shell launch commands * test(omp): consolidate shell fixture path import * preserve fresh OMP launch status * test: cover preserved OMP launch status * fix: recognize wrapped fresh OMP launches * chore(ci): refresh validation against fixed main baseline * fix(omp): recognize generated fresh launch guards across shells * fix(omp): preserve draft status and clear prefill across Unix shells * fix(omp): run cmd draft cleanup after either guard branch * fix(omp): launch drafts safely with nounset enabled * fix(omp): select draft shell without parser diagnostics * test(omp): await relay environment augmentation |
||
 Neilandstevelliu
|
ea02d90704 |
fix(omp): answer startup Kitty queries before renderer handoff (#20620)
* fix(omp): answer startup Kitty queries before renderer handoff Forward actual renderer capability through local and remote spawn. Preserve source ranges and following keyboard mode pushes, and retain independent ConPTY color authority. Refs #17081. Secondary review: #17082. Co-authored-by: stevelliu <stevelliu@tencent.com> * test(omp): cover fragmented keyboard modes and ConPTY handoff * fix: preserve keyboard startup intent without terminal colors * fix: negotiate keyboard support for host-authoritative agent launches * fix: keep terminal creation within line budget * fix(omp): negotiate keyboard support for paired web launches * test: remove obsolete message type import after main integration * fix: validate paired launch results and retry incomplete SSH test snapshots * fix(omp): negotiate keyboard support for background paired launches --------- Co-authored-by: stevelliu <stevelliu@tencent.com> |
||
|
NeilandBrennan Benson
|
4d3b0cca87 |
fix(omp): publish parent transcript paths for native chat (#20615)
* fix(omp): publish owning session transcript paths Adopt stablyai/orca#19529 on the child-session ownership fence. Preserve id-based resume and execution-host transcript boundaries. Co-authored-by: Brennan Benson <79079362+brennanb2025@users.noreply.github.com> * test(omp): render authoritative transcript reader output --------- Co-authored-by: Brennan Benson <79079362+brennanb2025@users.noreply.github.com> |
||
 NeilandNafisul Haque
|
f1a901e974 |
fix(native-chat): suggest verified OMP terminal commands (#20672)
Co-authored-by: Nafisul Haque <100821672+nafisul-haque@users.noreply.github.com> |
||
|
Neil
|
d966927013 |
fix(omp): submit large prompts in one PTY frame (#21573)
* fix(omp): join prompt submit with large paste * test(omp): cover joined submit timing * ci: rerun PR checks after timing test fix * test(omp): acknowledge joined submit activity |
||
|
c34b944136 |
feat(github): bind projects to a specific gh account (#13664)
* feat(github): bind projects to a specific gh account Adds per-project `Repo.ghAccount` so repo-scoped gh calls (create-worktree issue/PR search, work items, hosted-review reads and mutations) run as the bound account via ephemeral child-env token injection instead of the globally active gh login. Multi-account resolution is capability-gated (gh >= 2.40) and fails closed when the bound account or host is unavailable; Project View stays ambient by design. Repository settings gains a section for selecting or clearing a keyring-backed account (shadcn `Select`), with mixed-version "not enforced" handling for older remote runtimes. Attached `-Rhost/owner/repo` forms are covered by the host-drift guard and its tests; es/ja/ko/zh catalogs carry the section's strings. `getLocalProjectGhExecOptions` centralizes the binding lookup so every gh execution path picks it up, including the Electron `hostedReview:*` handlers that previously stayed on the ambient login. `gh auth token` (a keyring read) is exempt from the rate-limit breaker gate so a tripped bucket cannot turn a bound-token resolve into a false "unavailable". The `ghAccount` update field and the two binding RPC methods live in the shared RPC params contract; the generated catalog is regenerated. Fixes #13612 Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012B3QEP5iP4WGGEpPLtkHqA * fix(settings): make GitHub account refresh secondary * fix(github): satisfy strict casting quality checks * test(rpc): use runtime fixture for repo binding * fix(github): preserve project account for PR worktree lookups * test(rpc): avoid incomplete runtime settings fixture * fix(i18n): add GitHub account refresh label * fix(i18n): refresh runtime required catalog * fix(windows): preserve mobile patch bytes --------- Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com> Co-authored-by: Neil <neil@stably.ai> |
||
|
Neil
|
2038376d8e |
fix(terminal): a park must not discard the only copy of a remote pane's scrollback (#21285)
* fix(terminal): keep a client copy of a parked remote pane's scrollback
A remote-runtime pty's bytes never transit the client's main process, so the pane's
xterm buffer is the only client-side copy. The ordinary cold-park unmounted that pane
without capturing it, licensed by TERMINAL_PAIRED_PARKING_RUNTIME_CAPABILITY — a static
build string that says nothing about whether the host retained this pty's buffer. On
reveal, a host that answers 'no-serializable-buffer' (or stays silent past the request
timeout) collapses to a null snapshot and the pane paints blank: tabs and splits survive,
the scrollback is gone.
Capture before every park, not only the retention-budget force-park, so the reveal has a
copy to replay when the host cannot answer. An unverifiable host answer is not proof the
pane was empty; keep the buffer, never discard it.
Adds ORCA_E2E_FORCE_REMOTE_TERMINAL_SNAPSHOT_UNAVAILABLE so an e2e can reproduce the
host-retains-nothing state, mirroring the existing forced-truncation lever.
* test(terminal): prove a parked remote pane survives a host that answers nothing
The oracle is a token the test types into the terminal before the park and the fixture
echoes back. Nothing replays stdin, so a respawned command cannot reproduce that line —
only the pre-park buffer can. An earlier argv marker passed vacuously for exactly that
reason.
The control ('host retains the buffer') is insensitive to the fix and fails if the harness
never parks, never reveals, or never echoed the token, so the regression case cannot be
green for a harness reason.
* refactor(terminal): validate the paired host terminal RPC shape instead of casting it
The merge-commit consistent-type-assertions gate flags every new `as`. Two were fixture
shapes that a type annotation states directly, and the third hid an unchecked RPC payload —
readCreatedTerminalTab now fails with the shape named rather than surfacing later as an
undefined surface id.
* fix(terminal): let a park capture survive an unhydrated repo catalog
Reading state.repos unguarded threw out of the cold-park effect whenever the catalog was
absent, which would break parking itself. Capture is best-effort evidence; an empty catalog
also fails open in shouldPreserveTerminalScrollbackBuffers, the safe direction for a park.
* docs(terminal): pin why the two unhydrated-catalog fallbacks point opposite ways
shouldPreserveTerminalScrollbackBuffers fails open toward 'remote' because a worktree wrongly
judged local parks with no copy at all. worktree-runtime-owner.ts resolves the same unhydrated
catalog to 'local', which is safe there and would be data loss here. A reader pattern-matching
'fail open' across the two gets one of them backwards.
* fix(terminal): keep a parked pane's scrollback across a reconnect merge
The direct-SSH pull replaces a replaced tab's layout wholesale, and a park capture does not
bump tab.generation — so a just-parked tab is not in locallyPreservedTabIds and the only
client-side copy of its remote scrollback went with the layout it replaced. That is the same
data loss this branch already fixes, one layer down, and it is the layer that decides whether
the fix survives the app update the user actually performed.
Carry the client's leaf-keyed scrollback into the host's layout, filtered to the host's own
root leaves. Structure stays the host's verbatim, so a split it added while we were away still
wins and a leaf it retired still drops its bytes. Local wins a conflict: neither copy is then
the only one, but remote-wins would overwrite the tail captured since the last upload and
propagate that backwards on the next replace-session patch.
Not a generation bump: the pane key is `${tab.id}-${tab.generation}`, so bumping would remount
the pane and destroy the very buffer the capture just serialized, lift the recovery-storm
ledger ceiling, and let a stale local ptyId win through preserveNewerLocalTerminalFields.
* fix(terminal): carry a parked pane's scrollback through the mirrored-layout rebuild
Found in review of this PR by rc-ssh-remoting. chooseRemoteTerminalLayout rebuilds a
mirrored tab's layout from the host's picture and never carried buffersByLeafId or
scrollbackRefsByLeafId forward, though it already receives existingLayout. The host
publishes no scrollback of its own, so ANY session-inventory frame landing between park and
reveal dropped the only client-side copy: the rebuild is bufferless, terminalLayoutEqual
compares buffers so the write is not bailed out, and apply-terminal-records assigns it
wholesale.
Measured before the fix: 336 bytes captured at park, 0 after one forced frame, blank pane on
reveal. After: 411 bytes survive the frame and the reveal repaints.
The e2e passed either way because no frame happened to land in its window, so it was not
covering the destroying event. It now forces one inside the park -> reveal window and asserts
the capture survives it.
An identical fix was written and reverted earlier in this branch as 'no measurable effect' —
that measurement ran on a harness deleting the client profile between launches, so nothing
downstream of persistence could register. It was never actually tested.
* feat(session): add a local-only home for ordinary-park scrollback
localOnlyScrollbackByTabId is a top-level session field, tabId -> leafId -> buffer, that never
rides the remote projection: exportRemoteWorkspaceSession is an explicit allowlist of named
top-level fields, so a new one is omitted for free, whereas anything added to
TerminalLayoutSnapshot is copied whole. It is also outside the two records the mirrored-tab apply
rewrites, so a host inventory frame cannot wipe it.
Registered in every exhaustive session registry ('tabKeyed'), hydrated and scoped like the layout
map, dropped with its tab on close/removal/purge/repo removal/mirrored retirement, copied on profile
transfer, emitted by the incremental patch builder, and capped by pruneLocalTerminalScrollbackBuffers
alongside the shared home — with a per-home test so an uncapped path cannot go unnoticed.
Known ceiling, not widened here: the field routes through the partition router that falls back to
'local' when the repo catalog is unknown at write time (#21295).
* fix(terminal): keep ordinary-park scrollback off the upload, and read both homes through one resolver
The ordinary cold park fires on every workspace hide. Its capture now splits: structure (root,
ptyIds, titles) stays in the shared layout, bytes go to localOnlyScrollbackByTabId. Force-park,
hibernate, sleep and shutdown keep writing buffersByLeafId, because that copy is what a second
desktop cold-restores from; a shared capture clears the local copy so the two homes never hold two
versions of one leaf.
resolveLeafScrollbackBuffers is the only read across the two homes (local wins a conflict: it is
the later write by construction). restoreTerminalPaneLayout no longer reads buffersByLeafId
directly, the capture's merge prior comes from the resolver, and the post-replay release covers
both homes.
Measured with the projection at 20 tabs x 2 panes at the per-leaf cap: the shared-layout shape
exports ~22 MiB per replace-session; the local-only shape exports the bufferless baseline.
* test(sync): pin that the mirrored rebuild carries the client scrollback refs
The carry-through added in
|
||
|
Jinwoo Hong
|
ac4dc6599b |
feat(mobile): the desktop lists a page route, the shell honours it or stays native (OTA phase C, C1.3) (#21502)
* feat(mobile): the page mounts on the shell's init, with the client injected (OTA phase C, C1.1) The Route A entry built no client and mounted the route tree immediately, so the web provider minted its own: it read the page channel, built `BridgeRpcClient` and fell back to a placeholder that rejected every call. A tree that mounts before `init` reads synchronous getters against a client that knows no host, no state and no build, and the first render it records is the wrong one. The entry now owns the page's one client. It builds it from the channel at module scope, mounts nothing until `onReady` fires, and stamps the session and build ids `getShellSession()` returns on the document beside the mount state, so a screenshot, the render check and a device console read the same three facts. `client-context.web.tsx` takes that client by injection and serves it from `acquire()` for every hostId, because the bridge protocol names no host; the placeholder and its `BridgeTransportUnavailableError` are gone, along with the entry that pointed at them in the unvalidated-port inventory. A document with no channel is not inside the shell, so it says `unbridged` and stops rather than waiting out a backoff nobody answers. The render check gains a shell double that answers `ready` with `init`, reads the stamped session back off the document, and proves the gate is real by opening the same route with no double and finding an empty `#root`. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * feat(mobile): the shell names the screen, and the page routes to it (OTA phase C, C1.2) The shell serves its document at `/` and refuses every other path, so the page's own location matches no route in the tree it carries and expo-router paints Unmatched. Nothing in the document can tell it otherwise, so the screen has to cross the bridge. `init` gains an optional `route: { pathname, params }`. The pathname is held to what a path may be rather than to what a screen may want: rooted, single-slash, no query and no fragment. A protocol-relative `//host` would make `history.replaceState` throw a cross-origin SecurityError and take the mount down with it, and the params are a field of their own so neither side parses a URL. The shell route supplies it, the screen passes it to B4's hook, and the hook holds it for the life of one host: the page routes once, before its first render, so a route that changed afterwards has nothing left to change. The page writes that URL into its history and then mounts. It also hands the same URL to `ExpoRoot` as its `location`, because `ExpoRoot` snapshots `window.location.href` when its module is imported, which is before any frame has crossed the bridge: without it the router reads the `/` the shell served and replaces the page's own path right back. A shell too old to name a route leaves the page with nothing to open, so it paints a panel saying to update the app, built as elements outside React because the route tree is exactly what cannot mount there. Both platforms stop reading the document's URL to decide a load finished. The page rewrites its own path before its first render, so a document that committed at `/` reports finishing at `/h/<hostId>`; reading the path withheld `ready` forever and left the Android WebView hidden behind it. What is left is whether the load committed, which is the question the state machine already answers. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * feat(mobile): the desktop lists a page route, the shell honours it or stays native (OTA phase C, C1.3) The worktree list now renders from the desktop's bundle, and which routes do is negotiated rather than decided on one side. The manifest gains `routes: [{ pathname, grants }]`, written from one declared list the builder checks against the tree it bundled, so a declaration naming a screen with no module fails the build instead of reaching a phone as a page that paints Unmatched. The field is additive because the phone reads the manifest loosely and pins no schema version; the desktop's own writer stays `.strict()`, and the stale comment saying there was no additive path is corrected. The shell answers for what it can do. A route the bundle does not list, or lists needing a grant this app does not implement, settles as `native-route` and downloads nothing; so does a desktop that ships no bundle at all, which is the one blocked verdict that is not a wall, because a desktop with no bundle declares no page route and there is no workspace to refuse. The route is answered before the compat verdict for the same reason: a bundle this shell cannot open is not a reason to refuse a screen it was never going to open. `app/h/[hostId]/index.tsx` mounts the shell when the flag is on and takes the native list back as the fallback, and both routes read the flag through one hook so the census stays the whole census. A tap on a worktree row still opens the native session screen. The page posts `notify { name: 'navigate', href }` behind the `navigate` grant, which is not a convention: `notify` is a closed union, so an older shell refuses the whole frame and the page checks the grant before it posts. The shell pushes the target over the still-mounted view, so Back reveals the page with nothing reloaded. `route-handoff.ts` and its web sibling are the seam, router-shaped so the list's own hook and the recorder's adapter are untouched and no golden moves: the web file wraps the three members that leave the document and hands back any target outside the page routes `init` named. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * feat(mobile): the page can tell the shell it faulted (OTA phase C, C1.1) A page that throws where it renders has nowhere to report it: the shell sees a document that loaded and a view that never painted, so it waits on a blank page forever. This adds the one frame that says so. `notify { name: 'fault' }` carries the capture an `error` frame already carries, so both directions share one bound and one reader. It rides a grant because `notify` is a closed list on both sides: a page served by a newer desktop into an older shell would have the whole frame refused, so the page asks `init.grants.native` first and stays quiet on a no. The shell answers it as `document-load-failed`, which is what happened. That reason drops the generation and downloads once, so a page broken by bytes this host has since replaced recovers, and one broken by its own code stops at the failure screen rather than a blank one. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * refactor(mobile): give the bridge's notifications and the host's errors their own modules The fault report took both files over the 300-line cap, so each gives up the group that was already separable. The page's one-way members move to `bridge-client-notifications.ts`, which is also where the two policies that split them can be stated: the two the native contract declares throw before a session, and the fault report never throws at all. The host's three error classes move to `bridge-host-errors.ts`, the mirror of the page's own `bridge-client-errors.ts`. No behaviour changes. The commit before this one is over the cap on its own, which a forward-only history is the reason to say rather than hide. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * feat(mobile): one boundary under the page's root, reporting to the shell (OTA phase C, C1.1) Nothing in `app/h/**` exports an `ErrorBoundary` and `ExpoRoot` provides no global one, so a throw while a route renders — or a route module that rejects once the manifest is lazy — unmounts the tree and leaves a blank document. The shell sees a load that finished and waits on it forever. The entry now wraps what it mounts on `init` in one boundary that posts the throw over the bridge. Above `ExpoRoot`, not inside its wrapper: a route that cannot be resolved throws where the router renders it, and a boundary below the router never sees that. It renders nothing and offers nothing to press. The generation is on disk and was hash-checked before the view loaded it, so the same bytes throw again and a retry here would only throw twice; recovery belongs to the shell, which drops the generation on the report. The render check now grants the fault and collects what the page posts into the errors every case already asserts empty, because a throw the boundary caught paints nothing and logs nothing a `pageerror` listener would hear. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): write the page-fault callback ref after the commit, not during render React may replay or discard a render, so the write belongs in the commit phase. Layout, not passive, and declared above the host's effect: a native frame can arrive between a commit and a passive effect, and the host must already hold this render's callback. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): write the route ref after the commit, not during render Same class as the page-fault ref: render must stay pure because React can replay or discard it. Folded into the one commit-phase effect above the host's. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): write the page-route and navigate refs after the commit Same class again: the last two writes this branch adds join the commit-phase effect, so nothing this hook holds is written while React is rendering. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): take the boundary test to C0.5's fake-client pair `createBridgePortPair` is generic over the shell client now; the fake-client form this test wants is `createFakeBridgePortPair`. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * feat(mobile): bound the wait for a page that never says a word (OTA phase C, C1.1) A route module that throws while the bundle is evaluated takes the entry with it. The document still commits and the WebView still reports it loaded, but no boundary mounts, no fault is posted and no frame is ever sent, so the session sat in `ready` behind a blank view forever. The native view's finished load starts a clock; the page's first `ready` stops it; expiry is `document-load-failed`, which deletes the generation and fetches once. Nothing cancels the timer — a `ready` that lands first makes the expiry a no-op — so the runner owns a clock and the reducer owns every decision. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(config): make a route chunk throw, so the render check proves the boundary reports The check folded page faults into its errors but nothing ever produced one, so a boundary that stopped reporting would have stayed green. The server now serves one real route chunk with a throw in front of it: the module still links, so the failure is an evaluation throw where the router renders, which is exactly what the boundary is for. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): make the host enforce the grants it issued, and hear nothing before ready `forwardNotify` acted on any frame that parsed, including a `fault` from a page that had never asked for a session and therefore held no grant. Both refusals now go through one rule the host shares with the frame it sends, so the list a page is told about and the list it will be served cannot drift. Inert while every page is offered `fault`; the ungranted arm is what C1.3 needs the moment a grant belongs to a route rather than to the protocol. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): refuse a route no page can open, rather than blanking the WebView (OTA phase C, C1.2) `sendInit` put `options.route` straight on the wire and only the page's decoder checked it, so an out-of-contract pathname made the page refuse the whole `init`, ask again on its 2 s backoff forever, and the shell un-hide a view that would never paint. The only trace was a `console.warn` inside the WebView. Three changes, one failure mode. The host parses the route at construction and serves no session at all when it will not do, reporting it as a shell failure. The pathname rule refuses empty segments, dot segments and backslashes anywhere, because `replaceState` normalises `/../../etc` to `/etc` and `/h/a\b` to `/h/a/b` and the page then renders whatever came out. And the producer encodes the host id it interpolates, which is how one carrying a query, a fragment or whitespace got there. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): make a handoff mean the shell took it, not that a frame left (OTA phase C, C1.3) `handOff` returned `client.notifyNavigate(href)`, which answers whether the frame left the page and never whether the shell accepted it. Two hrefs the app builds today were posted, answered true and suppressed the local fallback, so the tap did nothing at all: the Connection-log link's object form, which `String` turns into `[object Object]`, and any href carrying a fragment, because the pathname is stripped to match and the whole href is what goes on the wire. Object hrefs now resolve the way the router resolves them, and the string is checked against the envelope's own pattern and cap before it is posted; anything that fails falls through to the local router, which is the policy this module already states. Whether a target names a screen that exists is shape's business no longer, and the comment says C1.7 owns it. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): start a new flow when the shell view remounts A remount cleared `pageReady` but left the flow alone, so the wait the retired document armed still matched. It expired onto the page that replaced it, took a ready workspace to `document-load-failed`, and deleted the generation on the way. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): say which page notification the bridge refused and why A refused `notify` fell through to the line about a view outliving its host, which is a different fault and names neither the notification nor the reason. The two refusals now get a line each, so a page that was told nothing cannot bury one reaching past what it was told. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): pin the ready deadline to the page's own retry ceiling The margin was stated in a comment and asserted against itself, so changing either number left the suite green. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): say what was wrong with the screen a refused shell named C1.1's per-kind log lands on a branch that also refuses a route, and that diagnostic was still falling through to the line about a view outliving its host. It names the shell's own bug now, and carries the issue. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): refuse a dot segment however the route spells it A URL parser percent-decodes a path before it resolves it, so `/h/%2e%2e/x` climbed out of the `/h/` prefix exactly as `/h/../x` does and landed the page on a screen nobody asked for, with no refusal anywhere. The one segment rule both patterns share now reads the encoded spellings as the dot segments they are, and still lets an escape inside a name through. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): name routes in the manifest field list the builder emits C1.3 added `routes` to every manifest this builder writes, and the Phase A contract test still listed eight keys, which is what went red in CI. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): hold a navigate target to the same segment rule as the shell's The href pattern is built from the segment source C1.2 tightened, and nothing said so: a spelling one pattern refused while the other took it would be a hole with a `notify` already pointed at it. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): give the ref-refresh probe the navigations this branch added C1.1's new case builds its own probe, and on this branch a probe also collects the hrefs the page hands back. The file stopped typechecking on the merge, which the tests ratchet caught. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * style(mobile): format the web shell route entry Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): split the bridge frame suite along the modules the merge created `bridge-rpc-client-frames.test.ts` reached 835 counted lines once C0.8 and C1.1 both added cases to it, over the 800 the lint allows. The split follows the two modules those changes extracted, so each suite now names the module it covers. `bridge client page faults` moves to `bridge-client-notifications.test.ts` (the outbound notify surface) and `bridge client refusals and send failures` to `bridge-client-inbound-frames.test.ts` (the reader, including the refused-event release that cancels at the shell). The seven suites that exercise the client as a whole stay put. The fake port all three drive moves to `bridge-page-client-test-harness.ts` rather than being copied three times. No case changed and none was dropped: 48 `it` cases before, 37 + 4 + 7 after, and all nine `describe` bodies compare byte-identical to their originals. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): type the shared init fixture as the member a case reads The harness exported `INIT` as `BridgeHostMessage`. While it was a module-local const, control flow narrowed it to the `init` member at each use, so `INIT.grants` read fine. An imported binding keeps its declared type instead, so the same read lost `grants` to the union and the tests ratchet went red. Declared as the init member, which is what every case already treats it as. No cast: the object literal is checked against the narrower type directly. `INIT` was the only exported fixture with this shape. `CONNECTION` is `as const`, `GRANTS` is inferred, and nothing reads a member off an `eventFrame` result. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb |
||
|
Neil
|
cd81725d70 |
feat(terminal): configure URL click and middle-click behavior (#21438)
* feat(terminal): configure URL click behavior * fix(i18n): include terminal link setting title * fix(i18n): localize terminal click controls * fix(settings): update terminal URL click title |
||
   
|
09073086a8 |
feat(terminal): inline images via @xterm/addon-image (perf-first) (#19512)
* feat(terminal): inline images via @xterm/addon-image, perf-first Add opt-in inline terminal images (SIXEL, iTerm2 IIP, Kitty graphics) through @xterm/addon-image, designed to keep idle terminals unaffected. Performance: - The addon (base64-inlined wasm decoders + protocol handlers) loads off the boot critical path via a deferred loader that mirrors the WebGL addon: primed after first paint only when the setting is on, read back synchronously at attach, with a 3-attempt cap so a transient failure never disables images for the session and a missing chunk never refetches per pane. renderer-boot-graph guards against eager import. - enableSizeReports:false so the addon never sets windowOptions and double-answers Orca's own CSI 14t/16t responder. - Perf-tuned decode/storage limits (storageLimit, sixel/iip/kitty size caps) in one place. Correctness: - Orca's DA1 handler wins over the addon's (last-registered-first), and the default DA1 response never advertised Sixel (;4), so DA1-detecting tools (chafa, img2sixel, viu, timg) never emitted it. The winning handler now appends ;4 while the setting is on, resolved per query so a live toggle changes the next DA1; idempotent against the ConPTY response that already lists it. - ORCA_IMAGE_PROTOCOL=kitty is exported to spawned shells (local, daemon, relay/SSH) and forwarded across the WSL boundary, so image-capable agents can pick an encoder. Unknown image sequences are swallowed by xterm when the addon is detached, so this never garbles output. - Settings toggle (default on) gates rendering and DA1 advertisement. Cross-checked against community PRs #7775, #11706, and #19201 at the end; credited below. Co-authored-by: s546126 <s546126@users.noreply.github.com> Co-authored-by: XRX193 <XRX193@users.noreply.github.com> Co-authored-by: lmsh7 <lmsh7@users.noreply.github.com> * fix(terminal): bound inline image memory and classify Kitty replies * fix(terminal): bound image decode and release image resources on cleanup * fix(terminal): address image addon review feedback * test(terminal): stub setPaneInlineImagesEnabled in appearance manager fakes * fix(terminal): evict unplaced kitty payloads before displayed images Byte-budget eviction dropped the oldest transmitted blob regardless of placement, so a new upload could erase a visible image while abandoned blobs still held budget. Unplaced payloads now go first and displayed ones only when that is not enough. The incoming image is always stored, so an oversized one overshoots the cap by one payload instead of being dropped after the protocol already acked OK. * fix(terminal): gate DA1 Sixel on real addon attachment; claim SSH image spec in CI - DA1 advertised Sixel from the setting alone, so a pane whose lazy addon chunk was still loading (or had failed all three attempts) told feature-detecting tools to emit DCS that nothing could render. Track the attached decoder per terminal and require it before setting the ;4 bit. - tests/e2e/terminal-inline-images-ssh.spec.ts was Docker-gated but claimed by no lane runner, so pr-e2e-gate-contract failed and the spec would have self-skipped green forever. - Reject non-positive PNG IHDR dimensions before decode: they are parsed with signed shifts, so a dimension >= 0x80000000 came back negative and slipped past the pixel-limit comparison. - One resolveTerminalInlineImagesEnabled() for the default-on setting; the four call sites mixed '?? true' with '!== false', which disagree on null. - One readInlineImageResources() walk of the addon internals instead of two copies that could drift against the patched dependency. - Isolate the deferred-attach drain per pane; make the zoom-invariance and backing-storage e2e assertions fail when the feature is dead. * refactor(terminal): one lazy xterm addon loader for webgl and image terminal-image-addon-loader was a structural clone of the webgl one — same memo, attempt cap, and .then(ok,err)-clears-memo recovery. Both now wrap createLazyXtermAddonLoader; each keeps its literal import() specifier so the bundler still splits the chunk (verified against a fresh build: addon-image stays out of the boot graph). * refactor(terminal): name openTerminal's addon flags; pin image addon limits Two adjacent optional booleans could be swapped without a type error once inline images added the second one. * docs(terminal): state the real per-pane image ceiling; drop test ordering dependency storageLimit:32 reads like the pane's budget but keys three pools — decoded pixels, retained encoded Kitty blobs, and pending WASM decoders — so the worst case is ~98 MB per pane with no cross-pane governor. Say so at the constant. pane-inline-images.test.ts's deferred case needed to run first; it now takes a fresh module instead, and the rest prime in beforeAll. Verified by running the file with that test moved last. * fix(terminal): satisfy rebased static analysis gate * fix(terminal): complete casting gate cleanup * fix(terminal): recover failed image addon loads * fix(terminal): bound image decoder allocations --------- Co-authored-by: m4air <m4air@m4airs-MacBook-Air.local> Co-authored-by: s546126 <s546126@users.noreply.github.com> Co-authored-by: XRX193 <XRX193@users.noreply.github.com> Co-authored-by: lmsh7 <lmsh7@users.noreply.github.com> Co-authored-by: Neil <4138956+nwparker@users.noreply.github.com> Co-authored-by: Neil <neil@stably.ai> |
||
|
Jinwoo Hong
|
73a58bd21a |
feat(session-search): resolve Workspace and Project scope on the host (#21509)
* refactor(session-search): move the AI Vault project key to shared The host must spell a project key exactly as the client does, so the two sides share one function instead of two copies that can drift. * feat(session-search): add a scope identity to the search request The panel cannot keep translating a project into one path per worktree: a repo with 580 of them exceeds the 64-path cap and the search fails outright. The request now carries the scope's identity instead, and a host acknowledges the scope it resolved so a client can tell a scoped answer from an old host's unscoped one. * feat(session-search): resolve a scope identity on the host that answers Every entry point already funnels into searchSessionService, so the identity becomes paths there once: native, WSL, SSH and relay hosts cannot disagree. A host that does not know the workspace or project answers scope-unknown rather than widening the search to everything it has. * test(session-search): pin how a host resolves a scope identity Covers prior paths, a workspace another now claims, folder workspaces, a custom worktree base path, flat placement where the global root belongs to every project, and the 580-worktree fold the panel's path list could not do. * fix(session-search): type the scope store by what the catalog reads A full Repo/Project/ProjectHostSetup requirement forced test stores to stand up rows the catalog never looks at. * feat(session-search): send the scope identity from the panel Workspace and Project name what to narrow to; All sends nothing. A host that answers a scoped search without acknowledging it is reported as needing an update, and none of its hits are shown, because they are not this scope's. * test(session-search): pin the new-client-against-old-host skew An old host strips the identity and answers with every session it has, and the answer is well-formed. The missing acknowledgement is the only evidence, so the merge drops those hits and names the host instead. * test(session-search): pin the identity and acknowledgement across every entry point IPC, the runtime RPC method, the relay handler and the shared remote client each carry the identity out and the acknowledgement back, and the relay -- which has no repo catalog -- reports the scope rather than widening the search. * fix(session-search): acknowledge the scope on an all-computers merge The merge built its results without the acknowledgement, so the renderer read it as an old host, dropped every hit and asked for an update. That is the default path: the panel defaults to Workspace and the host scope falls back to All. Per-host skew is still reported through `hosts`. Host-resolved paths no longer travel in `filters.scopePaths`. That field is capped at 64 for the clients that write it by hand, and the scanner child re-parses the request with the same schema -- so a project whose worktrees do not share one managed directory failed at 65 paths with "not ready". They ride beside the request now, where no wire cap applies. Managed directories come from buildKnownOrcaWorkspaceLayouts, so a workspace root the user has since moved away from is covered too. A workspace identity is resolved through this host's own worktree registry rather than the directory embedded in the client-supplied id. * test(session-search): follow the service search signature Host-resolved paths are a second argument now, so the call-shape assertions that pinned a one-argument call name it. * fix(session-search): answer consent and readiness before an unknown scope The registry short-circuited an unresolvable scope before current.search ran, and current.search is where disabled and not-ready are decided. A host with indexing off that lacks the project told the user it did not have the workspace, which they cannot act on. The verdict now travels to the service beside the request, and the service answers it after its own checks. * fix(session-search): acknowledge only a scope that resolved An unknown verdict is still a verdict, and it was being acknowledged as if the host had narrowed. The skipped banner also counted only 'searched' as having resolved the scope, so a host that resolved it and came back stale or timed out let the scope lines reappear where they explain nothing. * refactor(session-search): drop the version-mismatch receipt No stable release ships search, so the only hosts that have it and predate `within` are dev and ad hoc builds. The acknowledgement, the needs-update outcome and the copy behind it would be permanent dead weight from the first stable release on. The scope-unknown outcome and the off / not-ready / unknown ordering stay. Also trims this PR's new docblocks to the repo's one-line why rule. |
||
|
Brennan Benson
|
66e0847398 |
fix(agent-status): stop an auto-reviewed Codex approval reading as "Needs You" (#21389)
* fix(agent-status): stop an auto-reviewed Codex approval reading as "Needs You" Codex runs its PermissionRequest hook as decider #1, ahead of both its own review agent and the user, so the event means "a decision is being made", not "a human is blocked". Under the "Approve for me" posture the review agent resolves it seconds later, so every gated tool call drove the pane from Working to Needs You and back, plus a desktop notification each time. The execution host now reads the turn's approvals_reviewer off the rollout it already tails for subagent reconciliation, and keeps a reviewer-owned approval as working. Positive evidence only: an absent field, an older rollout, or an unreadable file all still raise the wait, so this can never hide a real prompt. Splits the incremental rollout JSONL cursor out of the subagent transcript module, which the new reader pushed over the file-length cap. * fix(agent-status): avoid stale Codex approval ownership * fix(agent-status): reconcile Codex child approval ownership * perf(agent-status): avoid reads for Codex child activity * fix(agent-status): scope Codex reviewer ownership by transcript |
||
|
Neil
|
0d7381d1f2 |
fix(terminal): keep an unverifiable park-reveal snapshot apart from an empty pane (#21396)
On a park-reveal of a remote-runtime pty the host snapshot probe is the only
structural paint (the reattach carries no relay tail). Every non-snapshot
answer collapsed to null with no retry, so a host that stayed silent past the
request timeout, or answered 'no-serializable-buffer' ("not proof the pane is
empty"), painted the same blank pane as a host with nothing. That reads
unverifiable as exited (docs/reference/ssh-execution-boundary.md).
Classify the probe three ways: a host image paints; permanently-unavailable /
unavailable paints nothing and asks nothing; everything that proves nothing
(timeout, host declined for now, local lane gate, imageless success) paints
nothing and hands off to the hidden-output restore loop, which already budgets
retry-worthy answers (7 host declines / 30 local gates / 5 re-arm cycles),
repaints from the host on success, and ends in the explicit loss banner. The
reveal's own probe is charged to that same budget, so the bound is shared, not
doubled. No structural clear is issued on the unverifiable path, so whatever
the layout replay painted from the client's own copy stays visible.
|
||
|
Brennan Benson
|
9907117569 |
feat(native-chat): record an explicit provider outcome on every structured turn (#21278)
A structured turn that FAILED was recorded as `completed`, identically to one that succeeded, so nothing downstream could tell them apart. Claude mapped only its two abort reasons to `interrupted` and let an API error fall through to `completed`; Codex collapsed every non-`completed` status to `interrupted` and read a missing status as a clean finish. Add `outcome` — success / failure / cancellation — to the turn record, emitted by both providers. The four-arm lifecycle union is deliberately untouched: it stays a report on what the HOST observed, and its readers are unaffected by construction. Absent means UNKNOWN and never success. Historical rows, older hosts, and any end the host inferred rather than heard (the child going away, a turn superseded before its result) all carry no outcome, so a newer client cannot mistake an old host's `completed` API error for a clean turn. Claude's abort-reason list had a second copy in the provider-fallback reader; both now classify through one `claudeResultOutcome`, so the durable verdict and the visible error row cannot drift. |
||
|
Neil
|
4e3170a76e |
fix(accounts): free the account queue when a sign-in is abandoned, and show the Codex sign-in link (#21372)
* fix(accounts): free the account queue when a sign-in is abandoned Closing Settings mid sign-in left the `codex login` / `claude auth login` child running, and every account mutation shares one FIFO queue, so the next Add Account sat behind it for the login's whole deadline and then inherited the abandoned call's timeout toast. Cancel the pending login before enqueueing the next add or reauth (never inside the queue the abandoned login owns), give Codex the cancel handle and Cancel button Claude already had, and stop reporting a cancellation as a failure. Also surface the sign-in link Codex prints, with copy and open, so the flow can be finished in a private window or another browser profile. * test(accounts): drop the bare casts CI's changed-code gate rejects The service doubles still need a cast; one documented helper per file carries the SAFETY rationale instead of nine bare `as never`s. * fix(codex): a cancel must not discard a sign-in that already succeeded The Windows post-auth watcher gives a lingering codex login five seconds to exit after it writes auth.json. A cancel arriving in that window rejected the login, and the caller's rollback then deleted the managed home that had just authenticated. Refuse the cancel once new credential bytes exist: there is nothing left to cancel, and the close handler already treats that state as success. Found by review of #21372. * fix(codex): keep a refused cancel cancellable, and require the sign-in notice Review of the auth-aware cancel guard found two holes it opened: - The outer handle latched `cancelled` before asking the session, so a refusal killed cancellation for the rest of the deadline. On a host with no post-auth watcher that reinstated the very stall this PR removes. Latch only when the cancel is accepted. - WSL never reads a pre-spawn baseline, so the guard read the auth.json that was already there and refused from the first click, making a WSL reauthentication uncancellable. Require a baseline before refusing. Also from review: publish the sign-in link from a stdout-only buffer, so an interleaved stderr chunk cannot truncate it; require codex's own "navigate to this URL" notice rather than offering the first link in the output; hide the notice in a remote account scope, where it would name a login running on this desktop; and share the cancellation message instead of matching a duplicated literal. The Claude case joins the login-process suite that already owns the two neighbouring cancel cases, and the auth-snapshot helpers move out of the session file, which the additions pushed over the line cap. * refactor(codex): cut the sign-in-link plumbing to its smallest form Review found the change correct but larger than it needs to be: - The pending-link store was a class with one permanent subscriber, a never-called unsubscribe and a try/catch that could not fire. It is a field and a listener set on the service, beside the cancel handle it already owned — and the service now clears both in one place. - The optional login-session dependencies were always supplied. - The parser's https check could not fail; the pattern already fixed the scheme. The renderer's unmount guard inside a synchronous IPC listener could not fire either. - The broadcast channel and the cancellation message are single sources of truth in src/shared now, rather than exported next to a hardcoded copy of themselves. - The duplicated seven-line rationale in both services says the same thing in three, including why only add and reauthenticate supersede. - The codex suite reuses its own factory, and unmocks once. Also reverts four reformat hunks the formatter pulled in around edits. * fix(accounts): free the queue for a switch, not only for another add Switching or removing an account shares the mutation queue an abandoned sign-in was holding, so the commonest thing a user does after giving up — pick a different account — still spun for the whole deadline while Add recovered instantly. Both now supersede, as does the Claude side. Every caller is a person: the two IPC handlers and the mobile RPC methods. No poll, sync or CLI path reaches them, and a sign-in that already wrote credentials refuses the cancel, so a switch cannot discard one that succeeded. Also from review: the Cancel button regains the gap its Claude twin has (layout is allowed by the design-system rule; only the colour override was not), and the URL subscription says what it is — registration for the process's lifetime, with no teardown to hand back. |
||
|
Jinwoo Hong
|
9641a1b544 |
feat(mobile-web-bundle): serve the packaged mobile web bundle over RPC (OTA phase A, 3/5) (#21348)
* feat(mobile-web-bundle): serve the bundle manifest and chunks over RPC Two paired-runtime methods on the already-authenticated connection: `mobileWeb.bundle.manifest` returns this install's manifest plus the chunk size it advertises, and `mobileWeb.bundle.chunk` returns one aligned range of one asset with the whole asset's length and hash, so a single chunk describes what it belongs to. `path` is accepted only by exact match against a manifest member, so traversal is unreachable rather than mitigated. Each asset's on-disk sha256 is verified once and the verdict remembered, concurrent first readers sharing one hash. Reads are capped at four in flight per connection, and a disconnected client stops costing reads at the next checkpoint. No SSH or relay proxying: a runtime answers only out of its own install. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile-web-bundle): pin the three buildId serializers against each other The canonical serialization exists in the builder, the packaging guard, and the shared contract, because the two packaging scripts run on bare node before any build output exists and cannot import TypeScript. A divergence in any one would reject every honest bundle at packaging, or ship a bundle whose id the phone recomputes differently and re-downloads forever. Proved red by swapping the guard's code-unit sort for localeCompare: five of six cases fail. Exports the guard's serializer for the test; no packaging behaviour changes. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile-web-bundle): cover every error code and a multi-chunk paging round trip Against a synthetic bundle in a temp dir, because the real builder's largest asset is under one chunk and CI unit jobs never build out/mobile-web. The fixture's script spans three chunks, its stylesheet is exactly one, and one asset is empty, so paging, the eof boundary, and the zero-byte case are exercised rather than assumed. Reads in flight are held by latching `open`, so the four-per-connection cap and an abort arriving mid-read are deterministic rather than a race with a stopwatch. Both were proved red: dropping the abort check after verification fails the abort case, and keying the cap on connectionId alone fails the device-token case. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile-web-bundle): resolve the bundle root through the AppEnvironment port check:runtime-electron-ratchet caught this: the resolver sat beside getBundledWebClientRoot in src/main/startup and imported electron, and importing it from an RPC method pulled the first electron edge into a runtime graph whose baseline is zero. The runtime has to stay bootable on plain Node. So it reads app.getAppPath() through the port every other runtime module already uses, and moves next to its two callers under src/main/runtime. A host with no environment installed has no install root, which is the same answer as having no bundle. orcad answers getAppPath from its own install root, so a headless runtime that carries the artifact serves it with no special case. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile-web-bundle): cover the resolver's two probe layouts directly Also stops exporting the manifest filename, which nothing outside the resolver needs. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile-web-bundle): pin both methods on the mobile allowlist The scanner only checks mobile-used ⊆ allowlist, and no mobile source calls these until A5, so deleting both entries left every test green. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile-web-bundle): keep filesystem failures inside the six error codes An asset unlinked or truncated after its verdict was cached reached the client as runtime_error carrying the desktop's absolute install path. Both now answer mobile_web_bundle_asset_changed, with the cause warned host-side only. A short positional read is the truncation case, so it throws instead of paging the client past the end. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * refactor(mobile-web-bundle): drop the unreachable release-idempotence guard The one caller releases exactly once in a finally; removing the flag left every test green, so it was defensiveness against a caller that does not exist. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile-web-bundle): prove a failed verify is not cached as a verdict The verdict cache never invalidates, so a transient read failure remembered as a verdict would poison the asset for the life of the process. Removing the delete left every test green until now. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * refactor(mobile-web-bundle): delete the unsatisfiable manifest params schema The dispatcher substitutes `{}` for absent params, so `z.null()` could never parse; the method declares `params: null` instead. A comment on the method name records why there is no schema. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile-web-bundle): fill the read window instead of failing a partial read fs.read may answer short of what it was asked for before EOF, so the previous check turned a legitimate partial read into a spurious asset_changed. The loop mirrors the relay's readFullStreamChunk, which is not imported because it sits behind the relay dispatcher's module graph; only a read returning nothing is treated as truncation. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * refactor(mobile-web-bundle): read the disconnect idiom with the shared predicate isClientDisconnectedError already exports exactly the check the catch needed, so the local error class goes away and the throw returns to the repo-wide idiom. The module doc now says asContractError is a total catch. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile-web-bundle): pin the four branches no test was holding Each one survived a mutation: the abort check before verification, the per-process manifest cache, the buildId component of the verdict key, and delete-at-zero in the admission map. The last two matter beyond hygiene — a verdict keyed by path alone carries a failed verdict onto the next build of index.html, and a map that never drops a key retains one pairing token per socket. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb |
||
|
Neil
|
a61119ceb0 |
refactor(runtime): name the four answers a host probe can give (#21207)
The renderer expressed every non-answer as one nullable `status`, so a probe in flight, a probe that failed, a host that refused us and a retired pairing all reached readers as the same `null` -- and readers spent that `null` on decisions of very different weight, including destructive ones. `RuntimeHostContact` names the four. Nothing changes yet: the connection-state derivation is rewritten on top of it and a 384-case parity table asserts the result is identical to a frozen copy of the old one on every combination of verification, transport, retired, answered and remote-control state. |
||
|
Neil
|
78a17bb24d |
fix(relay): one malformed pre-auth handshake frame closes its connection, not the daemon (#19879)
* fix(relay): one malformed pre-auth handshake frame closes its connection, not the daemon
parseHandshakeMessage returned whatever JSON.parse produced, and the daemon
interpolates the peer's version into a log line before any credential check.
A version that is an object with a non-callable toString throws TypeError
there, inside the frame-decoder callback. FrameDecoder.drainTurn wrapped its
synchronous dispatch in try/finally with no catch, so the throw escaped
feed(), escaped the socket data handler, and reached uncaughtException: the
relay daemon exited and every PTY and agent session it held died with it.
Two layers, because only the second closes the class:
- parseHandshakeMessage now requires the string fields each arm carries
(version; expected/got) and rejects a non-object payload. Both readers
share the parser, so neither side can interpolate a non-string again.
- FrameDecoder contains a frame owner that throws on the synchronous turn
the same way it already contained one on a continuation turn: reset the
residue and report one FrameDecoderContinuationError to onError. Every
owner's onError already closes its own connection, so any future throw
of this shape costs one connection instead of the process.
The relay CLI channel gains an explicit onError so a malformed reply still
ends that one-shot command instead of parking it.
* fix(relay): keep the diagnostic the refusal path exists to produce
Two error paths that destroy their own evidence.
`parseHandshakeMessage`'s unknown-type refusal interpolated `String(t)` on a
peer-supplied value: `{"type":{"toString":1}}` makes String() throw "Cannot
convert object to primitive value", so the refusal arrives without naming what
was refused. `describeRelayProtocolVersion` guards this exact hazard two files
away; the sibling was missed.
`runRelayOrcaCliChannel`'s new `onDecodeError` wrote to stderr and then exited
synchronously. stderr is async on a pipe transport, so the one line recording
why the command died could be dropped — the reason relay-handshake.ts already
exits inside its write callback.
* fix(relay): prove the optional handshake field too, not just the required ones
The parser refuses a non-string `version`, `expected` and `got`, then returns the
object with `endpointCredential` unproved — the most pre-auth field on the frame.
It is safe today only by accident: its one reader compares it, and a non-string
loses that comparison. Nothing holds that shape in place, and the next reader to
put it in a log line reinstates the template-literal throw this function exists
to stop.
Present-but-not-a-string is now refused at the parser. Absent stays absent: a
bridge presenting no credential is the common case, and refusing it would close
every unauthenticated-endpoint connection.
Wire-visible delta, deliberate: a peer sending a non-string credential used to get
`orca-relay-handshake-credential-mismatch` and exit 43; it now gets a bare close.
No first-party client can reach it — `runConnectHandshake` types the parameter
`string` and omits it when falsy — and a bare close is the right answer to a frame
that was malformed before any credential was checked.
* fix(relay): carry the SAFETY: rationale main's casting gate now requires
Main gained a `typescript/consistent-type-assertions` scan while this branch sat 432
commits behind, so every `as` the branch touches lands as a new finding. The parser is
the one place the handshake shape is proved, so each cast names the check that earns it,
and the hostile-frame cast in the round-trip test names the fact that it is a deliberate
lie the type system cannot describe.
* test(relay): annotate the hostile handshake frame instead of suppressing a cast
JSON.parse answers `any`, so a typed const expresses the same deliberate lie the
assertion did and the casting gate has nothing to flag. One fewer suppression.
|
||
|
Jinwoo Hong
|
3de77340fc |
fix: apply managed Claude auth to Agent Teams (#21356)
* fix: apply managed Claude auth to agent teams * test: update agent teams auth launch expectation * refactor: derive agent teams auth deletions |
||
|
|
d04b05b5c8 |
Detach retained CI and terminal tails from oversized strings (#20960)
* fix(memory): detach retained CI and terminal tails from oversized strings * fix(terminal): detach retained error and reattach string slices * fix(terminal): release oversized recent-output backing strings * fix(terminal): release backing strings held by PTY detectors * fix(memory): own bounded Claude background task labels * fix: detach retained terminal mode scan tails * fix: own retained plugin worker output strings * fix: own incomplete OSC 133 carry strings --------- Co-authored-by: m4air <m4air@Mac.localdomain> Co-authored-by: m4air <m4air@m4airs-MacBook-Air.local> |
||
|
OrcaWinandm4air
|
f90370fb6b |
fix: detach aborted shared auth filesystem waits (#21135)
Co-authored-by: m4air <m4air@Mac.localdomain> |
||
|
OrcaWinandm4air
|
0e3acf577d |
fix: release consumed runtime RPC queue entries (#21131)
Co-authored-by: m4air <m4air@Mac.localdomain> |
||
|
Jinwoo Hong
|
9c92136009 |
fix(relay): probe a half-open control socket instead of waiting out the silence bound (STA-7672) (#21076)
* fix(relay): probe a half-open control socket instead of waiting out the silence bound (STA-7672)
A socket CLOSE rejects a pending request as relay_control_closed_<code>, so a
relay_control_request_timeout is positive proof the socket stayed open and
simply never replied. The only thing that reaps such a socket is
RELAY_CONTROL_SILENCE_LIMIT_MS = 75_000, combed every 15s, against a 10s
request deadline. On Windows behind NAT/VPN or across sleep-resume a half-open
TCP socket accepts send() into a dead pipe and stays invisible for 75-90s, so
every pairing attempt in that window times out. The reporter burned ~7.
A request that times out with no inbound frame since its send now arms an RFC
6455 ping probe. Terminating on the timeout alone was rejected: relay control
ops run DB transactions that can outlive the deadline, and the existing comment
in handleMessage records that self-closing on a late reply was strictly worse
than ignoring it -- it orphaned the relay session and answered the phone with
HOST_OFFLINE for minutes. The probe distinguishes the two cases instead of
guessing.
The probe deadline deliberately exceeds the relay's own 15s application-level
ping cadence. Relay liveness never depended on RFC 6455 control frames
surviving end to end, so a shorter window would let a middlebox that swallows
pongs turn every request timeout into a reconnect loop. At 20s a healthy cell
clears the probe either way -- with a pong, or with the ping it was going to
send anyway -- so a probe that fires means the pipe carried neither. Detection
drops from 75-90s to ~30s.
A pong clears a probe but deliberately does not feed the silence watchdog: it
proves the pipe, not that the relay still indexes the session.
The timeout error also stops being a bare string; it now names the request
kind, the cell, the socket age, the time since the last inbound frame, and
whether a probe was armed.
The silence watchdog, the probe and the socket age now live in one
RelayControlLiveness owner rather than scattered across RelayControlClient.
* fix(relay): require a run of unanswered probes before tearing down a control
A single unanswered probe was treated as proof of a dead pipe. STA-3320 already
established that it is not: a cellular/VPN blackhole or a stalled TCP
retransmit routinely swallows one pong from a peer that is still there, which
is why RemoteRuntimeServerHeartbeat requires three consecutive misses. The
networks this detection exists for are exactly the ones that drop a lone frame,
so the first cut was more trigger-happy than the rest of the product.
Three changes, all aimed at the cost of a false positive rather than the
detection itself:
- Three consecutive unanswered probes are now required. The interval drops to
8s so the full run (24s) still outlasts the relay's 15s application-level
ping, preserving the property that a healthy cell clears the probe even where
a middlebox swallows RFC 6455 control frames. Detection lands at ~34s rather
than ~30s, against 75-90s before the fix. Any inbound frame retires the whole
run, so a later probe never inherits an earlier miss.
- The deadline carries the fleet's existing +/-10% jitter
(RELAY_RENEWAL_JITTER_RATIO). Without it every host timing out against one
slow cell would probe and terminate on the same boundary -- the synchronized
cohort burst that constant was introduced for. The pre-existing 75s watchdog
comb has the same defect; this path does not add to it.
- A liveness teardown now names its cause in the log. It reaches the origin as
an ordinary 1006 close, so without a label a probe-driven reconnect is
indistinguishable from any other drop, and a fleet-wide false positive would
be invisible in exactly the incident where it matters.
Mutation-checked: a miss limit of 1 fails four tests, 2 fails one, and removing
the jitter fails one.
* fix(relay): keep the request-timeout rejection classifiable
The diagnostics added in the previous commit were appended to the rejection's
message, which silently destroyed the signal they were meant to add.
`mobileRelayMintFailureFromUnknown` classifies a relay failure by testing
`error.message` against an anchored `/^relay_[a-z0-9_]{1,74}$/`, so
`relay_control_request_timeout reqKind=invite cell=...` stopped matching and
every pairing timeout was reported as the generic `relay_mint_failed` instead --
in exactly the flow STA-7672 is about. The pairing path logs only the resolved
code and discards the rejection's text, so nothing ever surfaced the suffix:
the change was a net loss of diagnosis.
The message is bare again and the diagnostics are logged from
RelayControlLiveness, which is the only place they survive.
Added relay-control-timeout-classification.test.ts to pin the contract end to
end through the real classifier, since the coupling is invisible at both sites:
restoring the suffix turns the assertion into relay_mint_failed.
Found in adversarial review.
* refactor(relay): collapse the half-open detection onto one object
Design review of the three commits on this branch. No behaviour change: the
184 relay tests pass unmodified, and reverting PROBE_MISS_LIMIT to 1 or 2, or
dropping the jitter, still fails them.
Dead plumbing. `probeIntervalMs` had zero callers across three layers
(client options -> conditional spread -> liveness default), and `silenceLimitMs`
the same -- the only production construction site, relay-control-origin.ts,
passes neither. Both are gone. `livenessRandom` stays; one test uses it. The
conditional-spread idiom went with them: `exactOptionalPropertyTypes` is off for
src/ (only cloud/apps/relay-ops sets it), so it bought nothing that
`?? Math.random` does not already do.
Teardown owns its own log. A two-member reason union crossed a module boundary
just to reach a console.warn, and the client re-derived `cell=` from
relayOrigin when liveness already held `cellUrl`. Liveness now tears itself
down and calls `terminate`; the client lost the import, the method, and the
exported type.
One probe object, one interval. `probeTimer` + `missedProbes` are now
`probe: { timer, misses } | null`, so "no timer implies no misses" is structural
instead of maintained by resetting in two places, and the
sendProbe/onProbeUnanswered mutual recursion is a plain setInterval. Jitter is
computed once per run rather than per tick -- one offset already desynchronizes
the cohort.
Honest probe label. If ping() throws, the old arm path returned false and the
caller logged `probe=in-flight/0` moments after terminating the socket -- a
false statement in the line that exists for incident forensics. The arm path
now returns the label it means, including `probe=send-failed`.
Absorbed RelayControlSilenceWatchdog. It had one consumer and no test file, and
this branch had to punch a `lastInboundTime` getter through it purely so
liveness could read state it holds. `lastInboundAt` now sits next to `openedAt`;
the file, the getter, the import, and the onDead('silence-limit') lambda are all
gone.
Also: dropped `RelayControlRequestTimeout.reqId` and `PendingRequest.sentAt`
(both written, never read -- the timeout closure captures the local `sentAt`);
dropped the two `'n/a'` branches, unreachable because a request timeout can only
fire after sendActive succeeded, which requires a state only handleProofMessage
reaches on the line before it calls liveness.start(); moved the classifier
invariant off a void-returning callback type and onto REQUEST_TIMEOUT_CODE,
where an edit to the string is next to the warning about editing the string;
and replaced the `live` parameter with an `isLive()` option so liveness asks
rather than being told, which also let `liveness` be constructed before
`requests` instead of a closure reading a field assigned on a later line.
|
||
|
Jinwoo Hong
|
c45b2c94c6 |
fix: make worktree scan failures actionable (#21291)
* fix: make worktree scan failures actionable * fix: preserve remote worktree scan diagnostics |
||
|
Jinwoo Hong
|
f2ca3cbfb7 |
feat(mobile-web-bundle): manifest and RPC contract for the desktop-served mobile web bundle (OTA phase A, 1/5) (#21325)
* feat(mobile-web-bundle): add the manifest contract and content-addressed build id The schema every later Phase A lane parses against: the ceilings that bound host memory (256 assets, 32 MiB total, 10 MiB per asset), and a build id that is a pure function of content so a client can use it as a cache key unconditionally. The serializer sorts its input rather than trusting the caller, so a producer that emits assets in any order still lands on the same id. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * feat(mobile-web-bundle): add the bundle RPC payload contract Method names, capability name, the 48 KiB chunk size, params/result schemas for both methods, and the six error codes as a closed union pinned by a coverage record. Constants and data only; the host wiring and the capability push land in later lanes. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile-web-bundle): hash the build id without node:crypto Metro ships no Node core shims, so a value import from these modules would fail to bundle on the phone. The pure-JS sha256 keeps both contract modules runtime-neutral, which also lets a cached manifest be re-verified on device. Verified digest parity against node:crypto across the 55/56/64-byte padding boundaries before the swap. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile-web-bundle): reject a manifest whose buildId is not its content hash A stale id passed every other check and would then serve the wrong bytes under a cache key the client already trusts. Runs last of the invariants because it is the only one that hashes. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile-web-bundle): require a lowercase content type The pattern carried an `i` flag over lowercase character classes, so the same bytes described as `Text/HTML` and `text/html` produced two different build ids. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * refactor(mobile-web-bundle): move the capability name to a zod-free module A4 wires this constant into protocol-version.ts, which the phone reads on the capability path. Leaving it in the schema module would have dragged zod along with it. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * refactor(mobile-web-bundle): name the chunk reply's length assetByteLength It is the whole asset's length, not the chunk's, and sitting beside dataBase64 under the old name it read as the chunk's. Both are non-negative integers, so a producer that emitted the wrong one would only surface at the final hash check. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile-web-bundle): reject asset paths that are not portable or that fold together Two paths differing only in case are one file on macOS and Windows, so the host would serve the same bytes under two entries and one of the two hashes could never match. Windows-reserved segment names and trailing dots cannot be written to the bundle root at all. Both follow skill-package-manifest's checks, the folded-path Set and the reserved-segment pattern. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile-web-bundle): accept one spelling of a parameterised content type The optional space in `; ?charset=` let the same bytes carry two content types and therefore two build ids. Pinned to the single-space form the bundle builder emits. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * perf(mobile-web-bundle): stop hashing a manifest a cheaper invariant already rejected zod runs superRefine even after the asset-array ceiling has failed, so a 257 asset manifest was still sorted and hashed. Each invariant now returns on its own issue and the count is checked first, which is what the comment claimed. The tests read the issue paths: an oversized or otherwise invalid manifest with a deliberately wrong buildId reports no buildId issue, while the same wrong id inside the ceiling does. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * docs(mobile-web-bundle): say that the manifest has no additive path `.strict()` plus a literal schemaVersion closes the shape completely, so the version bump is the only way to change it. The phone value-imports this schema, so Phase B must read an unrecognised schemaVersion as a bundle to re-fetch rather than as a parse crash. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb |
||
|
Brennan Benson
|
754134fd67 |
feat(agent-launch): deliver a launch prompt from the host (#21155)
* feat(agent-launch): deliver a launch prompt from the host `agent.launch` created the surface and then reported the caller's text as `not-delivered`, always: delivery lived in the renderer, so mobile and any other caller got an agent and no prompt. The host now commits a `submit` prompt to the structured session it just created, through the same send path `agentSession.send` runs, and reports `journaled` with the transcript row's id. Nothing is queued — the durable record that the text is owed is the journal's own submission row, which the send appends before dispatching, so a host-side copy could only disagree with it. The outbox's entry and envelope builders are reused so this send is shaped exactly like a client's, fingerprint included. Everything else under-claims as `not-delivered`: a terminal's paste is observed by whoever owns the pane, a `draft` has no host-side home, and a refused or thrown send commits nothing. There is no fourth "maybe" arm — a caller holding one could neither resend nor drop the text — and dispatch doubt stays on the submission row where it already lives. * fix(agent-launch): recover committed prompt after send errors |
||
|
Jinwoo Hong
|
40b2230508 |
test(mobile): typecheck the test files on a ratchet, and pin the reply enums where tsc looks (#21298)
* fix(mobile): move the last six reply-enum pins where tsc looks mobile/tsconfig.json excludes *.test.ts, so a `Record<HostUnion, true>` coverage record in a schema test is never typechecked: the two that existed (SshConnectionStatus, GitHubProjectOwnerType) checked nothing, and the four closed enums beside them had only a doc citation of the host type. Each arm list moves into its schema module as hostUnionArms<Union>(), which #21269 introduced for the same reason, and each test iterates the exported list instead of holding its own copy: - SSH_CONNECTION_STATUS to SshConnectionStatus - PROJECT_OWNER_TYPE to GitHubProjectOwnerType - DETAIL_FILE_STATUS to GitHubPRFile['status'] - PUSH_TEST_REFUSAL_REASONS and PUSH_REGISTER_REFUSAL_REASONS to the refusal arms of MobilePushTestResult and MobilePushRegisterResult - SETUP_RUN_POLICIES to SetupRunPolicy openEnum's parameter widens from a non-empty tuple to `readonly string[]` so a hostUnionArms list can feed it. z.enum already accepts the same, so the tuple constraint only excluded callers zod itself takes; behaviour unchanged. Twelve mutations prove the pins: dropping one arm and adding a bogus one each fail mobile tsc in all six places. Zero goldens move, the schemas' behaviour being unchanged, and the 21 recording suites pass at the existing baseline. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): fix the type errors in eighteen test files Found by typechecking the tests for the first time (see the config that follows). All mechanical, none weakens a product type: - 67 `act(() => vi.advanceTimersByTime(...))` callbacks return VitestUtils where act wants void, so each becomes a block. The async ones await only a genuinely promise-returning call, so no extra microtask tick is introduced. - Four fixtures were stale against a product type that gained a required member: MobileViewState.alwaysShowDefaultBranch, PrSidebarData.checksError, the branch-compare summary's errorMessage, and SessionOptionDescriptor's transport, which #20884 added precisely so a producer could not inherit the wrong lane's rendering by omission. - `getLastConnectedAt` on the shared relay fake was typed `() => null`, which refused the timestamp two escalation suites assign to it. - Two holders used before assignment take `!`, one `advance!.kind === ...` becomes `advance?.kind`, one widened status arm takes `as const`, and the Expo notification fixture keeps `data` required because the dismissal cases assign through it. 631 test files pass, 6222 tests, unchanged. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): typecheck the test files, on a ratchet mobile/tsconfig.json excludes *.test.ts so Metro never compiles tests into the release bundle, and vitest transpiles without checking types. Nothing had ever typechecked a mobile test, which is why a `Record<HostUnion, true>` pin written in one proved nothing and why 144 of the 630 test files had drifted. tsconfig.test.json is that program with the tests put back, behind `typecheck:tests`. Four files stay out: they import the desktop main process or src/shared/child-process, which are written against @types/node, and this program's libs are React Native's, where setTimeout answers a number rather than a NodeJS.Timeout. Pulling that graph in reports ~280 errors about the desktop rather than about mobile; vitest runs those four under Node, which is where they belong. The CI gate is a ratchet rather than the raw typecheck, modelled on check-ts-nocheck-ratchet.mjs: 126 files still fail, so the gate freezes that set and fails when a file that checks today stops checking, or when a baseline entry starts checking and was not pruned. The list may only shrink. Why not zero: 180 of the remaining 510 errors are one seam — tests locate mocked react-native components by string name, which `ElementType` does not admit — and closing it means either 180 casts or a global JSX declaration for the mocked names. That is a design decision, not a mechanical fix, so it is left for a follow-up rather than made here. The rest are smaller clusters of the same kind: vi.fn mocks assigned into typed slots, call-arg tuple indexing, and createElement props fixtures. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * docs(mobile-recorder): correct the corpus counts and the salvage claim The oracle section still quoted the corpus as 368 scenarios and 727 goldens; it is 393 and 778, and the three replay suites report 781 tests. Each number now names the command that measures it. "No golden carries one" was the load-bearing error: 44 goldens carry a recorded `reply-salvage` today, starting with the push-test unknown-reason scenario #21176 added for exactly that purpose. The paragraph claimed the observation pins an absence when on those families it pins a recorded drop. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): pin the tests-typecheck ratchet's parser The gate reads tsc's output, and tsc indents the "Overload 1 of 2, ..." detail under an error. Counting those as filenames would write unparseable entries into the baseline and leave the gate unprunable, so the parser is pinned on that shape as well as on the added/stale diff. Written against the gate itself: it flagged this file before the directive it carried was removed, which is the end-to-end proof the spawn half works. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): await the timer advances the act() rewrite dropped Rewriting `await act(async () => vi.advanceTimersByTimeAsync(n))` into a braced body left the returned promise floating at 27 sites, so the advance was no longer ordered before the assertions that follow it. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): unshadow MobileHostCard's .tsx suite A wildcard `include` keeps only the higher-priority extension, so MobileHostCard.test.tsx sat outside every tsc program while MobileHostCard.test.ts existed beside it. Its one error is the same react-test-renderer seam its sibling is baselined for. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): census every test file into the typecheck program The ratchet diffs only files that error, so a test excluded from tsconfig.test.json or shadowed by a sibling extension left the gate silently. Every *.test.ts(x) on disk must now be in the program or named in TESTS_OUTSIDE_PROGRAM with its reason. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(shared): make the enum helpers refuse the ways they can prove nothing openEnum takes a `const` T so a bare literal keeps its arms rather than widening to string. hostUnionArms blocks inference of U with NoInfer and defaults it to never, so a call that omits the host union — where the record would only pin itself — no longer compiles. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * docs(mobile): describe the census and correct the baseline count Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): give the push fixture cast its SAFETY rationale Widening the pre-existing cast made the changed-code gate attribute it as a new finding. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): build the push fixtures as typed notifications Replaces the `as unknown as` cast with Expo's own types, filling FirebaseRemoteMessage and its notification once in two builders, and passes the data payload in rather than mutating through an optional member. Typing the fixture showed one assertion comparing the scheduled content against the whole arriving content, which only held while the cast let the fixture omit the two members the presenter drops; it now names the four members the presenter forwards. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): keep the grouped-question advance read non-optional `advance?.kind` let an absent advance take the null-draft branch instead of failing. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * fix(mobile): run the tests-typecheck ratchet on Windows Spawns tsc's JS entry on this Node instead of the node_modules/.bin shim, which is a POSIX shell script that Windows resolves to tsc.CMD and then appends .exe to. Parsed paths are normalised to POSIX so a Windows run does not read every baseline entry as both stale and added. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb * test(mobile): close the ratchet's @ts-nocheck hole and read tsc once tsc exits 0 on a @ts-nocheck file, so a baselined test could be "fixed" with one line, pruned, and never checked again; the census now names any program test file whose leading comment carries the directive. `--noEmit --listFiles` answers both questions in one pass, so the gate spawns tsc once rather than twice. Corrects the two stale counts, and states hostUnionArms' real reason for living in the schema module now that tests are typechecked. Claude-Session: https://claude.ai/code/session_01JNnE9qzUZMMnqpZWCqM3nb |
||
|
Jinwoo Hong
|
4a86b2dc56 |
refactor(mobile): checked reply readers for files, dictation, host-screen and agent-history (step 7) (#21269)
* test(mobile): record main's file-preview and markdown-disk-fallback replies
Four of this branch's read sites had no malformed-reply coverage, so the reader
change would have had nothing to move at them. `familyGoldens` matrixes only the
first scenario of each family, and `files.preview-load`'s base is the grant-refresh
chain while `session.tab-documents`' is the served markdown tab — which left
`files.read` and `files.readPreview` on the worktree preview path, the artifact
image read, and the markdown tab's on-disk fallback recorded on their success path
only. This commit is the before picture, taken from main's own tree with no product
edit in it.
Three new families, five scenarios, ten goldens:
- `files.preview-worktree-text` / `files.preview-worktree-image` — `files.read` and
`files.readPreview` as the preview screen asks them for a worktree file.
- `files.preview-artifact-image` — `files.readTerminalArtifactPreview`.
- `session.markdown-disk-fallback` — the `files.read` leg a headless host's
`renderer_unavailable` sends the markdown tab down. It carries a second scenario
that serves `markdown.readTab`, because a matrix site needs a fulfilled reply
recorded somewhere in its own family to replay as the `normal` partition.
No existing scenario moved to a new family and no adapter changed, so every
pre-existing golden keeps its `adapterSha256` and `scenarioSha256`. Recorded in a
detached worktree at the manifest's pin (`4b876758d3`) with this manifest copied in;
the control is that all 748 pre-existing goldens came back byte-identical to
origin/main's, which `git diff
|
||
|
Brennan Benson
|
b66ef2e8a8 |
fix(agent-launch): resolve a launch scope, not a git worktree record (#21193)
* fix(agent-launch): resolve a launch scope, not a git worktree record `agent.launch` asked the runtime for a managed worktree record and then read exactly one field off it, `.id`. That record does not exist for every workspace a launch can run in, so the request refused launches the method could otherwise run: the floating workspace resolves to a scope with an id and a path but no worktree row, and `showManagedTerminalWorkspace` throws `selector_not_found` rather than hand back the id it had already resolved. A folder workspace survived that only because the resolver fabricates a worktree row for it. The scope is the answer that is real for all three kinds, so the launch asks for that instead. `showManagedTerminalWorkspace` is unchanged - callers that genuinely need the git record still get it, and still get the refusal. With floating now reaching the mode decision, the host must know which kind of workspace it resolved. The kind is derived from the id it resolved itself, never accepted from a caller, and the route module's existing `floating` blocker does the rest: a workspace with nowhere to keep a session runs a terminal agent. Behaviour change, deliberate: a floating-workspace `agent.launch` used to fail with `selector_not_found` and now succeeds as a terminal agent. That is what lets the floating titlebar agent button move onto the shared launch command instead of driving tab startup itself. No wire change: `AgentLaunchTarget` is untouched. * test(agent-launch): cover floating RPC workspace resolution |
||
|
Brennan Benson
|
434365d2de |
Offer to reconnect native chats that were working when Orca restarted (#21096)
* feat(native-chat): resume structured chats that were working at restart
Teardown records a marker for every session this host was genuinely running a
turn for, derived from the LIVE runtime rather than a persisted status row, so
a stale `running` row left by an older crash can never trigger a resume. On the
next launch a modal lists exactly which chats would resume and resumes them via
native continuation (Claude resume/resumeSessionAt, Codex thread id) — never by
re-sending the prompt, which is what makes an agent redo finished work.
A session resumes only when all of these hold: a teardown marker exists and has
not expired, the record's lease is released and reconciled, a provider resume
cursor exists and still matches the marker, the journal's own turn record names
the same turn, and the marker has not already been spent. Markers are consumed
before the resume is submitted, so a crash mid-resume cannot double-fire, and an
admission gate refuses a second concurrent resume for one session. Resumes are
staggered three at a time rather than spawning every provider at once.
The modal's "Don't ask again" checkbox writes the nativeChatResumeWorkOnRestart
setting, which Settings can turn back off; automatic mode runs the identical
predicate and staggering and reports what it did. Declining consumes the markers
so the prompt cannot return every launch — nothing is lost, because opening a
chat still re-acquires it at the same cursor.
* fix(native-chat): compare handle ROOT and turn state when offering a resume
Four defects QA found in the restart-resume offer, fixed together because the
first two interact: shipping the root fix without the state fix would convert a
silent no-op into actively offering finished chats.
1. Claude was never offered (0/4). The marker recorded agentSessionProviderHandleKey,
which embeds Claude's leaf uuid — a branch cursor. The adapter's own close path
appends a `resumed` link with an advanced leaf during the SAME teardown, so the
marker went stale seconds after it was written and the drift guard refused every
Claude session forever. Record and compare agentSessionProviderHandleRoot instead:
the root is the part a resume must preserve, and changing it is a fork, which is
exactly what this guard is for. Codex is unaffected (its thread id is the whole
key) but uses the root too, so the rule is uniform.
2. The predicate compared turn IDENTITY but discarded turn STATE, so a `completed`
turn satisfied it as readily as an interrupted one. Eviction rewrites `running`
to `interrupted` and never to `completed`, so the state is what separates work
that was cut off from work that finished. Require `interrupted` or `unverifiable`.
3. A chat blocked on a pending approval or question was marked as working, because
the teardown reader accepted any `running` turn while the product's own projection
calls that state `attention`. Teardown now defers to that projection: an agent
waiting on the USER is not interrupted work.
4. "Resume all" could silently no-op. The modal fetched candidates at mount; by click
time the chat's own pane may have bound and taken the hold, moving the lease to
`live` so the predicate dropped it and the call returned no results, leaving the
dialog open behind a dead button. Re-derive at click time and settle an
already-live session as resumed — it is running, which is what the user asked for.
Test fakes now model the Claude close path that advances the leaf, which is why no
unit test could previously exhibit defect 1. Ablation covers all eleven guards.
* fix(native-chat): gate the already-live settlement on the full resume predicate
Two follow-ups from re-QA, both cases of a rule stated by intent rather than by
discriminator.
1. The already-live path bypassed the predicate. "Resume all" sends no session
ids, so the fallback's target set was every marker, and it was gated only on
the session having a live provider child. A chat the predicate had refused --
a completed turn, say -- whose pane happened to own the lease was therefore
settled as `already_live` and had its marker spent, inflating the "Resumed N"
count with chats that were never eligible. No provider spawned and no tokens
were spent, but a marker the predicate rejected must never be consumed.
The resumable set now takes an explicit `leaseState`. The already-live path
derives a second set with ONLY the released-lease clause relaxed, and settles
a session just when it is in that set. Every other clause still applies.
2. The `attention` rule was one-sided. Teardown refuses to mint a marker for a
chat blocked on the user, but the set predicate had no equivalent, so a marker
arriving by any other route was offered once eviction rewrote its turn to
`interrupted` -- the same asymmetry the completed-turn case had.
Gated on projectStructuredAgentSessionStatus === 'attention'. That projection
tests for a pending approval or question BEFORE it looks at turn state, so it
still reports `attention` after the turn is settled, which makes it the durable
signal and keeps one source of truth with teardown.
Ablation now covers thirteen guards, including one for each of the above.
* fix(native-chat): capture awaits-user on the marker instead of re-deriving it
The awaits-user clause could never fire. It asked the live projection for
`attention`, which needs a prompt whose resolution is still `pending` -- but
teardown CANCELS that prompt a few phases after it writes the marker. By the next
launch the evidence is gone, for precisely the sessions the clause was written
for. QA measured the injection still being offered and then resumed.
This is the same shape as the leaf-drift bug: state read after teardown is not the
state that justified the marker. The discriminator, now applied across the whole
predicate:
- a fact teardown itself destroys or mutates must be CAPTURED on the marker
while it is still true;
- a fact that evolves on its own must be RE-DERIVED at read time, never
snapshotted.
So `awaitsUser` is now recorded at teardown and the predicate reads the recorded
value. Teardown still declines to mint a marker for such a session, so the
recorded flag is the second line rather than the only one.
Audit of every other clause against the same test:
- turn id (captured) -- teardown rewrites turn STATE but never the id. Correct.
- provider handle root (captured) -- the close path appends a resumed link, and
appendAgentSessionProviderHandleLink refuses one that changes the root, so the
root is invariant under exactly the mutation that broke the key. Correct.
- turn state (re-derived) -- DELIBERATE exception, stated here rather than left
implicit: we are not reading the state that justified the marker, we are
reading teardown's receipt that it settled the turn. A turn still `running`
means eviction never finished, and we refuse. Correct, and intentionally so.
- lease reconciled / released / handoff stage (re-derived) -- these answer a
different, launch-time question: may this host take the lease NOW. The
teardown-time value would be meaningless, and `unreconciled` is cleared by
this launch's own reconciliation. Correct.
- adapter support, marker TTL, marker consumption (re-derived) -- all evolve
independently of teardown. Correct.
Only awaitsUser was on the wrong side.
* fix(native-chat): drop the unreachable awaits-user marker flag
The captured flag was dead code. `awaitsUser` could only be true when the
projected status was `attention`, and `attention` hits the `continue` above the
push -- so every marker teardown can ever write carries `false` (QA measured
22 of 22 across two real teardowns). The predicate clause reading it was
unreachable by any production path.
A flag that is structurally always false is worse than no flag: it reads as a
safeguard, so the next person to touch this trusts it. The asymmetry it was
added to close was only ever reachable by fault injection, because teardown is
the sole writer of markers and already refuses attention sessions.
Removing it also drops an upgrade discontinuity: as a required field it made a
marker written by the previous build fail validation and be silently discarded,
costing a resume offer on precisely the upgrade where the user was mid-turn.
Markers predating the providerHandleRoot rename still will not parse, but those
carry a leaf-sensitive key the predicate would refuse anyway, so nothing usable
is lost.
In its place the teardown gate now states that `status !== 'working'` is the
SINGLE gate for awaiting-user sessions, why a predicate-side mirror would be
unreachable, and why it could not even re-derive the fact -- so the reasoning is
inherited rather than rediscovered.
Ablation is back to twelve guards; every other clause is unchanged.
* fix(native-chat): say reconnect, not resume, and show each offer's age
Two changes, both independent of the parked continuation decision.
1. The copy claimed something QA disproved. "Resuming continues each agent where
it left off" is false: reconnection restores the session at the point it
stopped, with full context and without re-sending the prompt, but the
interrupted reply does not continue on its own. The toast's "Resumed N chats"
implied work had restarted.
Audited every user-facing string against the rule that none may claim work
continues or that a reply resumes -- which caught more than the three strings
the fix started from. The title, the row button, "Resume all", "Resuming...",
the not-now hint ("picks it up where it left off"), the checkbox and its hint
("resume on their own"), the list's aria-label and the Settings row all made
the same claim. The user-facing verb is now reconnect throughout; the body and
update variant state outright that the interrupted reply will not continue.
en.json synced, runtime boot catalog regenerated.
If we later decide to send a continuation instruction, this is one commit to
change back. Shipping text we know to be false was the worse option.
2. Rows now show each offer's age. The TTL is 24 hours and a stale offer looked
identical to a fresh one. The marker already carried `recordedAt`, so this is
a render change plus one field on the renderer's candidate type, formatted
with the existing formatUiRelativeTime helper rather than a new one.
The clock is stamped once when the list arrives rather than read during render:
ages then stay stable across re-renders, and the render stays pure, which the
react(purity) rule requires.
Guards, predicate and RPC are untouched; ablation still covers twelve.
* feat(native-chat): show the workspace name on each reconnect row
A row read `codex · folder:8f3a1c22-… · 8 hours ago`. Recognising which chats
would reconnect is the entire point of the list, and at twenty rows a UUID
identifies nothing.
No RPC or host change was needed: the renderer can already resolve this id.
Resolved the way automation dispatch resolves the same id space
(resolveAutomationDispatchWorkspace) -- a folder workspace by its full
`folder:<uuid>` key via getKnownWorktreeById, a git worktree by its bare
`repoId::path` id via allWorktrees. Both return a Worktree, whose displayName is
a required field, and DetectedWorktree extends Worktree so either shape answers.
Falls back to the id when nothing resolves, which is what the row showed before
and also covers the window before the worktree store has hydrated.
The lookup lives in a per-row subcomponent because a hook cannot run inside
`map`, and its selector returns a primitive string so repeated selector runs
cannot churn referential equality.
* feat(native-chat): group the reconnect modal by worktree and add opt-in continuation
Grouping. Rows are now grouped under a worktree heading with the repo glyph and
an agent count, using the sidebar's own collapse mechanics. Only presentational
pieces are reused -- RepoIconGlyph, CompactAgentExpansion, AgentIcon and
formatShortTimeAgo. The sidebar's agent row cannot be: worktree-card-compact-agent-row
imports DashboardAgentRow, the dashboard's own type, so both surfaces render one
live-agent model requiring a pane, tab and status entry. Every chat offered here
is by definition stopped, so supplying that would mean inventing live state.
Two things I had assumed were reusable and were not:
- DashboardHostBadge returns null unless hostKind is ssh or remote. Structured
chat is local-only, so it would always render nothing. The host line is
omitted rather than faked; the badge is the right element to add if and when
structured chat gains remote support.
- No state dot. Every AgentDotState misleads here: idle and unverifiable both
presuppose a live pane, interrupted renders red like an error, done green,
working a spinner. A missing dot beats one saying these agents are running.
One worktree renders flat with no heading -- a name, count and chevron around a
single group says nothing the dialog has not already said.
The age column now uses formatShortTimeAgo for sidebar consistency. It takes
(timestamp, now) and subtracts internally rather than taking a delta, so the call
is (recordedAt, listedAt); passing the old delta would have rendered plausible
nonsense. The clock is still stamped once into state, so ages stay stable and the
render stays pure.
Continuation. A secondary "Reconnect and continue" action sends one message, from
a single shared constant, identical for both providers. Reconnect is unchanged and
still sends nothing. An info popover quotes the literal message read from that
same constant, so what is shown cannot drift from what is sent.
Ablation now covers fourteen guards. Two are new: continuation only follows a
reconnect that actually happened, and -- inversely -- a send injected into the
reconnect path must turn the test red, since "don't ask again" rests on reconnect
never sending.
* feat(native-chat): say terminal sessions kept running, and clear the quality gate
The modal lists stopped chats with no way to tell that CLI agents are fine, and
the true state of the world is counterintuitive: the terminal sessions survived
the restart and the chats did not. One line now says so, next to the heading
where it frames the list rather than as a footnote at the bottom.
Wording follows the app's own vocabulary rather than inventing a term: the
catalog settles on "terminal sessions" (terminalSessionCount, "Terminal sessions
are grouped by workspace", "No terminal sessions yet"), and UpdateCard already
reassures with "Your terminal sessions won't be interrupted during the update" in
the same text-xs text-muted-foreground treatment. "kept running" rather than
"were restored" -- nothing reconnected them, they never stopped, and the line
says nothing about why.
Also clears check:code-quality:changed, which I had not been running -- oxlint
alone covers neither the design-system nor the casting audit, so 18 findings had
accumulated across the branch.
- design system (4): Button spacing hand-rolled as gap-1/px-2 is just size="xs";
PopoverContent and DialogTitle own their typography and spacing, so the
text-xs moved to the popover's own children and the title's icon gap moved to
a plain wrapper.
- casting (14): production code loses its assertions outright via Reflect.get,
the idiom already used in managed-hook-detection-commands and
worktree-name-retirement. The marker validator reads each field through
Reflect.get and now checks recordedAt is a number rather than asserting it;
the store-file parse uses the existing `file` shape instead of a second
assertion; the runner narrows the admission error's owner with typeof.
Test fixtures keep their assertions behind the line-specific SAFETY:
rationale the repo mandates for exactly this case.
One trap worth recording: the audit reports an assertion at the line its
EXPRESSION OPENS, not where `as` appears, so a disable-next-line above the
closing brace of a multi-line literal is inert and silently changes nothing.
Guards unchanged; ablation re-proved 14/14 at this head.
* fix(native-chat): give the reconnect row's provider icon an accessible name
Every row rendered the provider as a bare AgentIcon, whose svg carries no
aria-label, title or alt. With a Claude chat and a Codex chat in one worktree the
two rows were identical to any non-visual consumer, and the dialog offered
several identically-named "Reconnect" buttons with nothing to tell them apart.
A regression from
|
||
|
Brennan Benson
|
7a1f55c52a |
fix(native-chat): give a failed Claude background task a typed row instead of an opcode (#20519)
* fix(native-chat): give a failed Claude background task a typed row instead of an opcode
A failed backgrounded command printed red rows whose visible text was the wire
opcode, and printed one failure twice. All five task lifecycle kinds are
catalogued status-chrome, but the payload sniffer in classifyProviderFrame runs
first and promotes any frame reporting a failure to the generic unknown-frame
fallback, whose sentence lookup has no key for the field Claude puts its own
sentence in. Two frames for one task therefore produced two rows, both of them
the method name.
Suppressing those frames is not the fix: when the last background task settles
the tracker flushes it and the strip unmounts, local_bash is excluded from the
subagent roster, and the status feed publishes only live tasks, so for a lone
backgrounded command the transcript row is the only report of the failure that
exists anywhere.
So the catalogue now binds: kinds a dedicated typed translator owns are named
as covered, and the generic fallback refuses to emit for them in either
direction. A new row owner keeps one durable row per task id, opened by the
announcement, revised in place by the lifecycle frames and closed by the
notification, carrying the provider's summary, error, output path, usage and a
run state. The row is written on the same dual carrier the subagent roster
uses: a frozen text twin plus a typed block, so a client without the block type
reads the sentence rather than nothing.
hasProviderError keeps its authority everywhere else, unchanged.
* fix(native-chat): keep tool attribution across a background-task row
A background task's row is a system message landing mid-turn between the
assistant's tool calls, exactly where the spawn-group roster row lands. Without
the same exemption it ended the run the following tool messages fold into, so a
tool result arriving after one stopped folding into its own assistant turn.
Also syncs the catalog with the row's one new translate key.
* fix(native-chat): harden background task rows
* fix(native-chat): settle background rows on provider end
* fix(native-chat): scope malformed task fallback text
* test(native-chat): assert only eligibility at the disposition layer
The malformed-task-frame test asserted the generic fallback resolves Claude's
`summary` field itself, which was true only while that key sat in the shared
key list. Eligibility is what this layer decides; the sentence the row leads
with is Claude's, supplied through the display-text seam and proven in the
translation test.
* fix(native-chat): gate background-task admission and scope rows per run
Admission now matches the reference on all three gates. Type is the whole gate
and MONITORS ARE NOT ADMITTED: a monitor runs for the life of the session and
has no outcome a row could report, so it never reaches the timeline. On first
admission only, the task's tool_use_id must name a tool call this session
forwarded at the TOP level — a Task spawned inside a subagent's sidechain names
an id that never reached the transcript, and a top-level row for it would claim
an invocation the user never saw. And a task that already exists and has not
finished is not re-opened: a duplicate announcement is a redelivery, not a
second run.
Rows are now keyed per RUN. A provider may reuse a task id for a distinct later
invocation, and a row keyed by the id alone overwrote the first run's transcript
history instead of leaving it standing. Generation 1 keeps the bare key, so
every row already written is unaffected.
The spawning tool call is carried on the row as parentToolUseId. Orca's journal
has no structural parent link for an item — AgentJournalItemIdentity has four
arms and none carries one — so the relationship is data on the item rather than
nesting.
A terminal frame that names NO tool still opens a row. That is a named
deviation, recorded at its call site, and the measurement behind it is in the PR.
* fix(native-chat): read the aggregate roster by membership, not a phantom status
The background-tasks payload types every entry as exactly
{task_id, task_type, description, ambient?}. It has no per-entry status, so the
state this owner derived from one was always undefined and the reopen branch it
guarded was unreachable on every real payload — proven by deriving the state
from an SDK-shaped entry and getting null.
Membership is the only liveness the payload carries: it is the whole live set
after a change, so presence means live and absence means merely "no longer
listed", never an outcome. Presence does not revive a settled row either — the
level's ordering against the start/stop edges is unspecified and it carries no
evidence of a new run, so the task's own frames stay the only thing that opens
or settles one. Only the identity fields it really sends are read, and ambient
housekeeping entries are excluded as the payload asks.
The two helpers that served the dead branch are removed, along with the test
that exercised it through a synthetic status the CLI cannot send.
* fix(native-chat): mirror reference task admission and drop the synthesis path
The forwarded-parent gate is conditional on the field being PRESENT. An
announcement naming a tool this session never forwarded is a nested child and is
refused; one naming no tool at all is admitted, because absence of the field is
not evidence of an unforwarded parent. The previous rule required the field and
so refused every tool-less task.
Terminal frames now match on task_id alone. The forwarded-parent question is
settled once, at admission, and is never re-asked on a notification or a patch.
A frame for a task that was never admitted yields no row, and a patch is folded
into the row it names rather than opening one.
That removes the synthesized-row path entirely, and with it the named deviation
it carried: the captured tool-less failure lands on a row that already exists,
because its own tool-less announcement is admitted. The dead builders go with
it.
Left deliberately stricter than the reference, and flagged rather than changed:
a terminal frame still records its task id as terminal even for a task never
admitted, so a late announcement cannot open a row for work already reported
finished. Two existing tests pin that.
* fix(native-chat): preserve background task ownership across restarts
* chore: restore pnpm-lock.yaml to origin/main
A local pnpm run rewrote the lockfile and the merge commit swept it in. The
branch changes no dependencies, so it must carry no lockfile delta at all.
* fix(claude): harden background task lifecycle
* fix(claude): bound task generation history
* fix(claude): preserve task identity after history eviction
* fix(claude): resolve background task identity after rebind
* fix(claude): isolate queued task runs
* refactor(claude): give the background-task ledgers one bounded owner
The bounded collections behind a background-task row were read out of the
class with `Reflect.get` to prove they stay capped, which the anti-slop
gate rejects. Move them into `ClaudeBackgroundTaskLedgers`, which owns
the caps beside the eviction helpers and reports a typed readonly size
view the tests assert against.
Also replace a `Reflect.get` in the mobile recording proxy with typed
property access.
* fix(native-chat): report a background task failure the transcript never admitted
A terminal `task_notification` for a task no announcement ever admitted rendered
nothing at all. The typed row owner declined the row because its map held no
entry for the id, and reported the frame as handled — which is exactly what
tells the generic provider-frame fallback to stay quiet. Both surfaces declined
the same frame, so a real failed background task was dropped on the floor.
A terminal frame is self-sufficient: it states an outcome, and it carries the
summary, status, error, output path and usage that outcome needs. It now opens
its own row from those fields, with the summary as the label and `unknown` as
the kind when the frame names no task type. The row map enriches a terminal
frame; it never gates one. Every terminal status writes one, not failures alone,
so there is one rule here rather than a third behaviour for failures.
The deliberate hand-offs still win, because they are recorded rather than
implied: ambient, subagent and foreground tasks are claimed in the foreign-owner
ledger the notification path already checks first. A Task spawned inside a
subagent's sidechain now records its refusal there too, under `sidechain`,
instead of leaving no trace and reading as a task nothing ever decided about. A
capacity-refused task whose outcome the generic fallback already printed records
`fallback` the same way, so a redelivery neither prints twice nor mints the row
capacity refused.
The anti-resurrection guard stays and stays scoped to announcements: a late
`task_started` cannot reopen work already reported finished. The restart rule is
now stated once instead of twice — a different parent alias is the provider's
restart signal only when BOTH runs name their parent, which is what the terminal
ledger already required of an evicted row and what the live row now requires too.
* test(native-chat): pin that an orphan task row reopens the provider's turn
* fix(native-chat): stop an orphan task row drawing its sentence twice
An orphan row took its header label from the notification's summary, which
also renders as the row's sentence, so the same string appeared in both slots
of the same collapsed row. The label now stays empty and the header falls back
to the task kind, leaving the sentence to carry the provider's words.
* fix(native-chat): keep Claude task outcomes owned through capacity and redelivery
* fix(native-chat): keep settled overflow notifications from reopening a turn
* fix(native-chat): retain Claude task rows through journal pressure
|
||
|
Brennan Benson
|
abc8386e14 |
fix(mobile): name a create's launch so a lost reply cannot build two workspaces (#21137)
* fix(mobile): name a create's launch so a lost reply cannot build two workspaces `agent.launch` admits a caller-supplied `operationId` through a durable ledger, so exactly one execution happens and every replay returns the recorded answer. No client sent one, so the machinery was inert and the original defect was still live: mobile retries a lost create by design, and a retried launch built a second agent in a second workspace. Mobile now mints an operation id per create candidate and sends it whenever the host advertises `agent.launch.replay.v1`. The invariant is one operation per candidate. `computeAgentLaunchFingerprint` folds `target` whole, so the workspace name is inside the fingerprint; carrying one id across a name-collision bump would meet its own row under a differing fingerprint and refuse `agent_session_operation_conflict`, failing the create outright on the second candidate. The id is therefore minted beside `clientMutationId` at the top of each loop iteration and reused verbatim by every retry arm inside that candidate — never re-minted, since a new id is a new operation. Admission runs ahead of every effect, so `_invalid` / `_expired` / `_capacity` prove nothing launched: those re-send the same candidate unnamed rather than let bookkeeping fail a create the host would have performed. `_unknown` is the one refusal that is not safe to re-send, and it surfaces. Also corrects a false comment: the legacy path caches the whole launch under `clientMutationId`, so inside its 60s window a replay adds neither a workspace nor a surface, and outside it adds both — not "a second surface, never a second workspace". * fix(mobile): preserve launch identity on refusals * fix(mobile): use launch receipts to authorize replay * test: move mobile launch replay coverage outside node project * fix(mobile): enforce replay-safe launch delivery at the host * test: run mobile launch contracts in mobile checks * test: cover mobile launch contract workflow dependencies |
||
|
Neil
|
96eb97aad6 |
fix(runtime): split the host-contact epoch out of the connection generation (#20359)
* fix(runtime): split the host-contact epoch out of the connection generation `connectionGeneration` carried two meanings and one reader was always wrong. holding the session mirror through an outage leaves its subscriptions stranded and that edge was the only thing left to revive them. But the same value is the mirror's cache key -- use-runtime-session-mirror-environment-key.ts keys the subscription effect on it, every published frame is stamped with it, and web-session-terminal-retirement-proof-ledger.ts drops retained proofs when it moves. So the bump #20085 needed as a resubscribe signal re-keyed and rebuilt the mirror after any brief flap, which is the #19647 symptom #19873/#20059 fix. Measured first: with the reconnect bump deleted, an ended stream followed by recovery issues zero resubscribes, and the mirror's subscribe call registers no `onClose`, so main's terminal close is dropped. #20085's claim is true -- the subscription really is dead after recovery -- so the trigger has to exist. It just must not be the cache key. Give each meaning its own value: - `connectionGeneration` returns to identity only: a new runtime session, a re-pair, an explicit clear. A same-runtime return no longer moves it, so no stamp, fence or retained proof is invalidated by a flap. - `hostContactEpoch` counts "the host answered again after we lost contact". It lives on the store entry and is read only as a dependency of the two subscription effects in use-web-session-tabs-sync.ts -- never passed to an installer, never part of `environmentKey`, so it cannot become a stamp. `useRuntimeSessionMirrorEnvironmentKey` becomes `useRuntimeSessionMirrorEnvironmentKeys`, returning `environmentKey` (identity) and `resubscribeSignal` (the epoch edge) from the one target scan, so the hot ownership scan is not doubled. Each direction is pinned by its own test: removing the resubscribe dependency fails only 'reinstalls both session-tabs subscriptions when the host answers again'; restoring the reconnect bump fails only the two key-stability tests. * test(runtime): pin the mirror hydration verdict across a host flap The generation tests assert the key string; this asserts what the user feels. The mirror's hydration verdict is stamped with the connection generation, so any bump discards it and every mirrored pane re-parks -- the tab-list rebuild. Held across an unverifiable probe, still discarded when the runtime id actually moved. * test(runtime): build real host statuses instead of casting partials |