## What changed
- Launch background compaction at eligible pre-turn, mid-turn, and post-turn checkpoints when `AsyncCompaction` is enabled, capturing the current world state and client routing state.
- Adopt ready results at compaction checkpoints while preserving newer messages, tool outputs, and world-state updates. Continue normally if an optional result is rejected.
- Run one foreground compaction if an adopted result still exceeds the context limit.
- Prevent cancelled turns from launching new background work while preserving already-running work for later reuse.
## Testing
Add unit and scenario coverage for stale history and account changes, cancellation, pending-result reuse and fallback, retained tool outputs, compaction event pairing, and token usage accounting. Add a context snapshot covering a collaboration-mode change while compaction runs.
GitOrigin-RevId: 04ac028d9fe2e3ea64aacaaeb351c3c776cf3940
## Why
Assistant messages delivered during a pending Guardian review can change its evidence. Previously, these context changes aborted the review instead of triggering a fresh assessment.
## What changed
Treat changes to local or root review context as stale authorization so Guardian retries with updated evidence. Cancellation and history resets still abort the review.
## Testing
Extend local and worker delivery tests to cover refreshed allow and deny decisions, verify that the new review includes the delivered message, and ensure commands execute exactly once after approval and never on stale approval or denial.
GitOrigin-RevId: 019205bee6d3f4e8cf6b9268f72bbc5db6b3899e
Set `config.model_catalog` to the bundled catalog in
`mcp_invalidation_refreshes_threads_that_are_still_starting` so model
resolution completes locally before the test waits for the startup
lifecycle hook.
GitOrigin-RevId: fd12df5140670dad4c9e2d22fabe73b0437af2b4
## Why
Cancelling an async migration caller can leave its blocking compression or decompression worker running. Releasing migration locks before that worker finishes allows other writers or migrations to access files still being processed.
## What changed
Share maintenance, cross-process writer, and local writer locks with blocking workers during migration and recovery. Keep those locks held until file work completes, releasing cross-process locks before allowing local writers to proceed.
## Testing
Add a regression test for cancelled compression and decompression callers that verifies locks remain held until the worker finishes and the output is complete. Preserve the Unix check that decompression temporaries have owner-only permissions.
GitOrigin-RevId: 74855250d89ecab81eaadc2956e328d37b65e72e
## Why
Keychain signing reconstructs its own authentication context, so invalidating the caller's context does not reliably cancel the context that owns the biometric prompt.
## What changed
Evaluate the stored key's access-control policy explicitly with `LAContext` before signing. On cancellation, invalidate that context and wait for its authorization callback before releasing the lifecycle lock, rejecting any late success. Reload the key with the authorized context and disable further interaction so signing cannot open another prompt.
## Testing
Update unit tests to cover waiting for cancellation completion and rejecting completed authorization after cancellation. Add an ignored macOS hardware test and setup instructions covering persisted Secure Enclave key reuse, signature verification, cancellation, retry signing, and cleanup.
GitOrigin-RevId: 4ea9756430c5649654e95e047d1666617503072c
## What changed
Add optional `voiceSessionId` to realtime start parameters and forward it as the `voice-session-id` header for WebSocket, WebRTC, and existing-call transports. Keep this client-generated voice session identifier distinct from `realtimeSessionId`, which uses `x-session-id`.
## Testing
Extend serialization tests and add coverage for distinct session ID headers, WebRTC request forwarding, and existing-call handshake forwarding.
GitOrigin-RevId: cb558dbb9d438f2d34e281199a7e81fe5d33b879
`nearestExistingParent` already checks whether the cache path exists and is a directory. Remove the duplicate `stat` check from `usableCacheDir` in the OpenAI docs manual fetcher, retaining the existing parent lookup and write/execute permission checks.
GitOrigin-RevId: 42583353bfdd06881171d8b24fa2b2d5b6fb3135
## What changed
- Return qualified skill names such as `team:deploy` and parsed `dependencies` and `policy` instead of a separate namespace and raw metadata file. Share `agents/openai.yaml` parsing and validation between skill loaders and executor discovery.
- Include each standalone skill's search-root ID and path in `discoveryRoot`, while accepting responses from older executors that omit it.
- Add plugin `displayName`, falling back to the manifest name or installation directory.
- Materialize inline MCP declarations in `mcpConfig` with the manifest as their source path, and report incomplete discovery when declared MCP or Apps configuration files have no returned contents.
## Testing
Add coverage for JSON round trips, omitted `discoveryRoot`, optional and invalid skill metadata, qualified names, display-name fallbacks, and inline MCP configuration. Verify standalone search roots survive cached discovery and sandboxed RPC responses.
GitOrigin-RevId: 4124972444ad7c2fb8c6418acbf9a589d797fad2
## What changed
- Accept `super` modifiers and normalize `cmd` and `command` aliases to `super`. Support capturing, displaying, and using these bindings in shortcuts, chords, and leaders.
- Add `super-left`, `super-right`, and `super-backspace` editor defaults for moving to the start or end of a line and deleting to its start.
- Honor safe Super-modified editor shortcuts during startup without treating them as text or inserting a pending paste newline.
- Save keymap edits through `config/value/write` when connected to a local daemon so it can validate bindings before they are saved or applied. Preserve the configuration and active keymap if the daemon rejects the edit.
## Testing
Add coverage for modifier aliases, shortcut capture and labels, chords and leaders, line editing, startup input, and daemon acceptance or rejection of keymap edits. Extend the startup PTY test with Command editing sequences.
GitOrigin-RevId: 8ed74f8253c89d423d5dfe87aa092844ce834454
## What changed
Apply `tui.copy_on_select` to composer mouse selections on release, including X11 PRIMARY publication when enabled. Keep editable selections highlighted so typing can replace the selected text, even when `tui.copy_on_select_clear_selection` is enabled.
Share clipboard destination handling across the transcript, footer, and composer. Apply automatic deselection to footer selections after confirmed delivery, and release composer and footer PRIMARY ownership on focus loss or resume. Update configuration descriptions to clarify the behavior across all three surfaces.
## Testing
Add composer tests covering copy on release, retained selection highlights, replacement by typing, PRIMARY ownership lifetime, and clipboard-only status reporting for pending copies.
GitOrigin-RevId: 46999a32170149193fca561741000e97dfc458cf
## Why
`codex app-server daemon restart` previously reused saved settings without applying feature flags supplied on the command line.
## What changed
Apply `--enable`, `--disable`, and boolean `-c features.*` overrides during an explicit restart and persist them for subsequent restarts. Accept nested `enabled` values, normalize feature aliases, and replace conflicting saved aliases while preserving unrelated settings.
Reject invalid feature toggles before stopping the daemon or saving settings.
## Testing
Add integration coverage for overrides on stopped and running daemons, alias replacement, preservation of saved settings, restarts without flags, and rejection of non-toggle keys without interrupting the running daemon.
GitOrigin-RevId: 14a657bc4758b9c248e363f9dac58f8887070a04
## What changed
When the remote global catalog is active, let a cached `visualize@openai-curated-remote` install supersede `visualize@openai-bundled`, matching the existing Sites behavior. A disabled remote install also suppresses the bundled plugin; a missing remote bundle preserves the bundled fallback.
## Testing
Extend the agent-turn compatibility tests to cover Visualize with enabled, disabled, and missing remote bundles, asserting which skills appear in the developer prompt.
GitOrigin-RevId: 7d0cc564189249e597d5bd29b5800353d89a9060
## What changed
Update the OpenAI docs sample resolver to return only `model`, `migrationGuideUrl`, and `promptingGuideUrl`, and remove the model-to-slug conversion helper.
GitOrigin-RevId: d7cea29ff8121b1a4df3b062d54dcbed6a1589bd
## What changed
- Add the disabled-by-default `executor_capability_discovery_v2` feature to acquire `capabilities/discoverV2` inventories from supporting remote executors using each selected environment's working directory and sandbox.
- Share snapshots with MCP and world-state contributors and skill providers without activating discovered plugins or publishing discovered skills to the model.
- Cache inventories by environment instance, executor session, working directory, and sandbox. Coalesce concurrent requests, refresh the cache at turn start, retry transient failures, and refresh MCP projections after recovery.
- Resolve the system skills root in the executor and pass it explicitly to capability location discovery.
## Testing
Add coverage for feature gating, unsupported executors, inert inventories, cache reuse and eviction, turn refreshes, failure recovery, executor replacement, and system skill locations.
GitOrigin-RevId: 1bdca28fa7f17622b27e2713ab3dcc0c13426889
## What changed
Use `conversation` when neither a local `async_classifier_mode` override nor a model default specifies a mode. Keep `snapshot` available as an explicit override and update the configuration documentation and schema to describe the new fallback.
## Testing
Expand configuration tests to cover all local override and model default combinations. Update app-server approval tests to exercise conversation mode without an explicit override, and explicitly select snapshot mode in tests that depend on it.
GitOrigin-RevId: a62479a3f6454bf28d589373a236d8b0ab5023b9
## Why
Compressed rollout searches recompiled the same query patterns for every file.
## What changed
Lazily compile and reuse the JSON-escaped and content search patterns across each scan, including the fallback scan when `ripgrep` is unavailable. Keep compilation failures within the existing per-file error handling and open files before compiling patterns.
## Testing
Add regression coverage verifying that an oversized query does not fail the compressed rollout scan and that a missing file still returns `NotFound` before query compilation.
GitOrigin-RevId: 0d41cbca6e1f716fc234c268493d8dd94c34975e
## Why
Opening the in-app resume picker created a new app-server session without the active session's remote working directory override, dropping the remote directory filter from the thread list.
## What changed
Pass the active session's remote working directory override to the picker session so its `thread/list` request retains the `cwd` filter.
## Testing
Add a regression test that opens the in-app picker against a recording remote server and verifies that `thread/list` includes the remote working directory in `cwd`.
GitOrigin-RevId: 0073ed77e9d9b0ffdd647fcb9f157d20ea2aa599
## What changed
Select label-only web link rendering when `TERM_PROGRAM` is `rex`, matched case-insensitively. Preserve the existing visible-destination fallback for multiplexers and `dumb`, `screen`, or `tmux` terminal modes.
## Testing
Add a regression test verifying that Rex renders `[label](https://example.com)` with the label-only policy.
GitOrigin-RevId: 615ca6186d84fc17cef9286e75848877b542b6ab
## What changed
- Register `capabilities/discoverV2` for initialized sessions and expose it through `ExecServerClient` and remote `Environment` APIs, checking executor support before calling it.
- Reuse startup discovery locations and retry failed prewarming on demand. Replay discovery once after executor recovery while preserving the request parameters.
- Track whether all required filesystem watches registered successfully and reload locations and metadata on each request when registration fails.
## Testing
Add coverage for RPC metadata refresh, omission of skill instruction bodies, sandbox isolation, startup scan recovery, disconnect replay, and successful or failed watch registration.
GitOrigin-RevId: 726e8f7e8b4d835b877f31d1d8a9f8eea1ad9c8b
## What changed
- Default `features.api_key_cyber_access_programs` to `true` and allow `codex exec` to select Daybreak with an OpenAI API key. Select the advertised standard program when `daybreak=false`, while respecting explicit feature opt-outs.
- Resolve API-key eligibility for TUI background turns from the target thread's feature settings, including project opt-outs. Omit access-program overrides when the server cannot report those settings.
- Preserve automatic server treatment for unlisted models and missing or empty Cyber program metadata in exec and TUI. Continue rejecting Daybreak when a nonempty program list does not offer it.
## Testing
Update regression coverage for API-key Daybreak selection, the off override, explicit opt-outs, resumed sessions, target-thread settings, and missing model metadata.
GitOrigin-RevId: f3cf7bd95a2754d4b4281bc70b18c7da4b177552
## Why
MCP servers must forward sandbox policy unchanged to `codex sandbox --sandbox-state-json`, while interpreting launcher information to find the Codex CLI. Keep those contracts separate so consumers can preserve unknown policy fields.
## What changed
- Accept the experimental `codex/sandbox-state` capability alongside the deprecated `codex/sandbox-state-meta` capability, and send identical policy values under both `_meta` keys.
- Move `codexExecutable` out of the policy into `codex/sandbox-launcher`, supplied only when a verified CLI path is available for a local stdio server.
- Replace stale sandbox metadata while retaining unrelated entries, and remove stale launcher metadata when no launcher is available.
- Document the metadata contract and how servers should forward policy and retain authority for long-lived subprocesses.
## Testing
Update unit and integration coverage for policy serialization, identical state aliases, launcher availability across execution environments, and stale metadata replacement.
GitOrigin-RevId: e544c06e6e023953b1fc8069b7b1affa065262d1
## Why
Rewriting source agent names and instruction filenames inside URLs and email addresses can change their destinations when importing external agent content.
## What changed
Use `linkify` to preserve recognized URLs and email addresses while rewriting surrounding text, including Markdown link labels and titles. Continue migrating instruction filename references to `AGENTS.md`, including relative links and references with line or column suffixes.
## Testing
Add rewrite coverage for URLs, email addresses, Markdown links, and surrounding Unicode text. Extend skill import and app-server configuration import tests to verify that link labels migrate while URL destinations remain unchanged.
GitOrigin-RevId: c4d210a287c99b6440438c99527364d379c568bd
## What changed
- Mark `cli_daybreak` as stable and enable it by default.
- Apply Daybreak to background turns only for eligible OpenAI accounts, allowing API-key accounts when `api_key_cyber_access_programs` is enabled. Preserve saved settings for ineligible accounts and report account lookup failures when Daybreak is enabled for the OpenAI provider.
- Skip inheriting Daybreak configuration when the app server does not support reading effective configuration.
## Testing
Expand dynamic-tool request tests to cover the feature toggle, ChatGPT and API-key authentication, API-key program eligibility, and unsupported configuration reads.
GitOrigin-RevId: 55a57315c9f527f977dd8305ea62d4096ed7dae9
## Why
Pending classifications can outlive the `max_tool_call_lag` window in which their scores are reusable, keeping metadata waits or classifier transports active after the result becomes stale.
## What changed
- Expire pending classifications when tool-call progress exceeds the configured lag allowance, preserving the wrapper discount while nested code-mode actions remain active.
- Cancel metadata waits and sampling in snapshot and conversation modes, including snapshot response drains after an early score.
- Treat expiry as supersession, preserving cached scores and the last completed conversation checkpoint and transcript cursor.
## Testing
Add regression coverage for lag boundaries, active wrapper discounts, transport cancellation before and after early scores, expired requests skipping preparation, and conversation history and transcript deltas surviving cancellation.
GitOrigin-RevId: 892ff13a05fd2b1573cef9842b934d1fc865d951
## Why
Canceled classifier requests need bounded cleanup that preserves received token usage and returns only safe WebSocket connections to the pool.
## What changed
- Own classifier response streams and connection leases together, interrupting dropped responses with a one-second cleanup deadline.
- Require a matching interruption acknowledgement and response completion before reusing a canceled response's socket. Retire sockets on rejection, invalid acknowledgements, or timeout while preserving any received completion and token usage.
- Skip requests canceled before dispatch and return their unused sockets to the pool.
- Distinguish cancellation from ordinary steering, which continues draining without requiring an interruption acknowledgement.
## Testing
Add regression coverage for cancellation, acknowledgement ordering and failures, cleanup timeouts, socket reuse, and token accounting under backpressure. Extend steering tests to cover rejected and missing acknowledgements.
GitOrigin-RevId: c4e7aac49d36cea2c2d0e9f2ba5f6147f67471dd
## What changed
Add transcript select-mode state with an underlined cursor, grapheme-aware movement via arrow keys or `h`/`j`/`k`/`l`, and vertical movement that preserves the preferred display column. Support marking with `v`, Space, or Shift-arrow keys and copying through the existing selection machinery.
Keep the displayed transcript snapshot and reading position paused during selection and after exiting the mode. Add entry and exit methods; this change does not yet bind mode entry to a user-facing key.
## Testing
Add unit tests for initial cursor placement, movement across empty lines and entry boundaries, wrapped text, preferred-column preservation, and Enter behavior with and without a marked range.
GitOrigin-RevId: e5bc380ea0eb6e2c6157ba73ac063d548a7de145
## Why
Encrypted compaction payloads can include sampled-token replay caches that increase their size roughly threefold. Budgeting them as ordinary text can reject checkpoints that fit the model context budget.
## What changed
Use a shared estimate of one token per twelve ciphertext bytes, rounded up, for encrypted `Compaction` and `ContextCompaction` items in history and Guardian context budgeting. Guardian continues to count framing and metadata separately.
Check parent compactions against the configured token budget and a separate hard wire-size limit. Preserve the complete encrypted checkpoint when it fits. The token estimate remains a heuristic and can underestimate legacy checkpoints without replay caches.
## Testing
Add coverage for token-estimate rounding, parent compaction token and wire-size limits, and reviewer continuation with a cached compaction preserved intact.
GitOrigin-RevId: cc18ca29802491865167eef992830efcf93d97da
## Why
Servers can advertise cyber access programs without accepting them for API-key accounts. The TUI must account for server support before exposing those programs.
## What changed
- Check `features.api_key_cyber_access_programs` at startup and when loading models for API-key accounts, clearing `available_access_programs` when forwarding is unavailable.
- Use thread-scoped feature settings after thread creation and honor feature requirements and CLI overrides at startup.
- Refresh account-scoped models when switching to or from API-key authentication or changing API-key thread context, clearing cached access programs while the refresh is pending.
## Testing
Extend session tests to cover enabled and disabled server settings, CLI overrides, and catalog filtering before and after thread creation. Update account-change tests to handle the model refresh.
GitOrigin-RevId: c80de30a0063b84949098d137715d8678bd70587
## Why
Prediction forks need to retain their parent's context even when it exceeds the automatic compaction threshold. Compaction or context resets can otherwise replace the history used for the prediction.
## What changed
- Apply a thread-local `AutoCompactionDisabled` policy to experimental prediction forks without changing model or skill budgets.
- Skip automatic compaction before, during, and after turns, including context-limit recovery, and suppress new token-budget reminders.
- Reject `new_context` requests for these threads and report context-limit failures without compacting their history.
## Testing
Add regression tests for legacy and paginated prediction forks across compaction thresholds, unchanged parent compaction and skill/tool inputs, context-limit failures, and rejected context resets. Add a request-history snapshot verifying inherited context and reminders survive without compaction or additional reminders after a tool call.
GitOrigin-RevId: 7df374287f4142f776981bb6c22e5caa564d74a5
## What changed
Replace `softprops/action-gh-release` in `.github/workflows/rust-release.yml` with explicit `gh` commands:
- Look up the release by tag, including drafts, and fail on API errors.
- Create stable releases as drafts until assets are uploaded. Create prereleases directly to preserve prerelease events.
- Upload sorted assets with `--clobber`, then publish the release with the configured title, notes, prerelease status, and latest status.
- Cap release notes at 124,999 characters, including a trailing newline.
GitOrigin-RevId: f343e4f265b06e047d85fe5a2b89a92690c7a7fb
## Why
A transient Apps MCP startup failure can remove cached tools that the model has already discovered, preventing a later Code Mode invocation from using them.
## What changed
Keep cached Apps tools in the model's tool catalog while startup recovery runs. Let invocation readiness start or join a shared retry before acquiring the tool execution lock, while tool discovery remains nonblocking. Keep recovery running if a waiting invocation is cancelled.
## Testing
Extend connection-manager tests to cover cached tool bindings and shared recovery. Add a Code Mode regression test that discovers a cached Calendar tool, invokes it after startup fails, and verifies that it succeeds after recovery while unrelated tools remain able to execute.
GitOrigin-RevId: e888501a1050884a50a473e71b893cf6672fe256
## Why
Terminal `response.incomplete` errors discard reported token usage. Counting that usage as a new context baseline can also double-count input and trigger unnecessary compaction on the next turn.
## What changed
- Carry the response ID and valid reported usage with incomplete-response errors.
- Record usage for sampling and foreground compaction in cumulative totals, rollout records, budgets, and token-usage contributors without replacing the context baseline. Preserve the original terminal error, including when accounting exhausts the budget or cancellation races the error.
- Charge incomplete background compaction requests to the rollout budget without publishing foreground token usage.
- Ignore missing or malformed usage rather than estimating it.
## Testing
Extend coverage for SSE and WebSocket incomplete responses, rollout usage records, compaction accounting, deferred token-count events, and token-usage contributors. Add regressions for unnecessary compaction on the next turn and background compaction publishing foreground usage.
GitOrigin-RevId: bf1ee6df3912052ab7c19e1c6af8ad3dc4b07675
## Why
The remote exec-server parent-lifetime test can hide an early process exit behind a connection timeout, making startup failures harder to diagnose.
## What changed
- Race the WebSocket connection against the child process exiting so startup exits are reported immediately.
- Include the registration request count and child status when attachment fails, without inspecting request headers or bodies containing authentication data.
- Enable remote exec-server info logging in the test while keeping other logs at error level.
GitOrigin-RevId: 75739fc93d7f28ba0a3f447a14d4d20b2dbdfd14
## What changed
Move child configuration, task path construction, fork usage hints, and backend spawning into `core::thread_spawn`. Add request, result, and error types that distinguish preparation failures from backend errors.
Route the v2 `spawn_agent` handler through the shared helper while retaining tool-specific telemetry, activity events, and response formatting in the handler.
GitOrigin-RevId: a09a8f66beb41c1403220750d236a12717dbfc6c
## Why
`with_executor_hooks` clones the hook registry to create an executor-specific view. Copying the configured handlers, plugin sources, and diagnostics on every clone is unnecessary when that configuration is unchanged.
## What changed
Store hook configuration in `Arc`-backed collections and use copy-on-write when replacing executor-scoped handlers. Skip handler mutation when there are no executor hooks to add or remove.
## Testing
Extend the executor hook registration test to verify that updating a cloned engine leaves the previous engine's handlers unchanged.
GitOrigin-RevId: 6b188ca6ca21772ccd7336344d98c023f769834e
## Why
Cancellation after initial input admission could leave a child registered even though the spawn caller had not received its result.
## What changed
Resolve the child's status and configuration before committing reservations and releasing provisional ownership. This keeps cancellation cleanup active through the final awaits and lets the spawn result return without yielding after commitment. Cancellation after the result returns does not retract the accepted child.
## Testing
Add a regression test that cancels spawn during the final status lookup after input admission and verifies that the child is unregistered, terminated, and unloaded, and that shutdown completes. Extend the provisional-child cancellation test to cover both native V1 and V2 callers.
GitOrigin-RevId: 028b873ad4c1fac2c4b5d1dc823b4561ff19e6bf
## What changed
Allow Guardian Decisions to use `OPENAI_API_KEY` when the default `openai` provider has no custom base URL and neither `CODEX_GUARDIAN_DECISIONS_API_KEY` nor a cached API key is available. Preserve cached API key precedence over the environment fallback.
Report missing credentials as `credentials_unavailable` instead of `provider_error` in Decisions failure metrics.
## Testing
Add tests for cached API key precedence and the environment fallback with ChatGPT authentication. Run each test in a child process to isolate environment changes.
GitOrigin-RevId: fbb772dcb9287d9e1be1b34292ad9bdffeb3d6b3
## Why
Non-parallel tool calls can acquire their lock in either order, making the shared board scenario's `last_message_id` snapshot nondeterministic.
## What changed
Send each `collaboration.post` call in a separate mocked response to order the writes before snapshotting. Update the `board_connections` snapshot and expected request count from 15 to 17 to reflect the additional responses.
GitOrigin-RevId: a290a01f5071a2081d4d0b1531b1b31bfbb8a6e6
Empty drafts need not survive a daemon restart. Update the `discovered`
scenario in `codex-rs/cli/tests/daemon_startup.rs` to rename the thread before
restarting the daemon and include `thread-name` in the status line. Wait for
the confirmed footer name before issuing `/daemon` so the test proceeds only
after the thread is materialized.
GitOrigin-RevId: a91508582826fa1bd0cc7b89d9a1af7e8b4e1999
## What changed
Emit `codex.guardian.review.truncation` counts and `codex.guardian.review.truncation.bytes` histograms after successful input finalization. Aggregate observations across reduction stages by component and disposition (`truncated` or `omitted`), recording original, retained, and omitted byte totals. Exclude feasibility checks and failed retries from loss telemetry.
Keep blocking command-hook cancellation fixtures in the local environment and use double quotes for the Python command.
## Testing
Add metric-export tests covering aggregation across multiple reductions, separate truncation and omission totals, no loss metrics for rejected inputs, and no duplicate reporting on repeated finalization.
GitOrigin-RevId: 69182fc7afe230be69dd2469caedb5ae4494947b
## Why
Remote process exit can arrive before the final output, and inherited pipes can keep the output stream open after exit. Removing the session at exit can lose output that should remain available to `write_stdin` polls.
## What changed
- Retain remote sessions with open or unread output after exit, including early exits and cancelled initial collection.
- Collect pending output when `write_stdin` races with exit, skipping stdin writes once remote exit is known.
- After initial collection finishes and the process exits, allow a five-second grace period before a final background read. Respect active polling deadlines and keep later polls available while that read recovers retained output.
- Let explicit termination stop output collection even when the process has already exited, and install the exit watcher before cancellable cleanup.
## Testing
Extend remote process event tests to cover output arriving after the initial yield, polling during a delayed final read, active polls exceeding the background grace period, resumable cancellation, and termination before the initial yield.
GitOrigin-RevId: b1671fad8b61a2cdecf7cf9fdf92e57c028c7af8
## What changed
- Await a compatible pending background request when foreground compaction is needed, and share the result installation path with foreground requests.
- Validate the captured history generation, prefix length, model, and authentication owner before installing a prepared result. Fall back to foreground compaction when required if the result is rejected.
- Replace only the captured history prefix, preserving later messages, retained tool outputs, goal edits, and the current world-state baseline. Record installation analytics and completion events only after a result is installed.
GitOrigin-RevId: 798963f2fc603c3febba2c21eafa1bcb094b3ebf
## What changed
Add `features.multi_agent_v2.disable_native_tools` to hide native multi-agent tools while keeping the V2 runtime and its instructions enabled for host-provided tools. The option defaults to `false`, preserving existing tool exposure.
## Testing
Extend configuration parsing tests for the new option and parameterize the multi-agent catalog scenario to cover both settings. Add a snapshot showing native tools are absent while the V2 runtime and instructions remain enabled.
GitOrigin-RevId: 21a53584d2395bb678813c876a8b7927bea61c90
## What changed
- Capture history, prompt metadata, routing state, authentication ownership, and context-window identity for background compaction requests.
- Send background requests over HTTP while preserving foreground routing, and retain the latest candidate in session memory.
- Add feature and threshold eligibility checks and deduplicate attempts by history version, model, compaction hash, and authentication ownership.
- Cancel and drain background tasks during session shutdown, preserving completed usage accounting.
The background launch helper is not yet connected to the turn loop.
GitOrigin-RevId: 0e5e6ab7dac4d98a836733668a775ac48b5a59f1
## What changed
- Extract an HTTP Responses executor that can run without owning a turn's WebSocket state, with optional authentication-owner checks.
- Add cancellation-aware compaction output collection that records background token usage against the rollout budget.
- Support installing a compacted history prefix while preserving later history items and placing initial context around real user or agent messages.
- Allow callers to supply a context window for history trimming and a captured window for compaction request metadata.
GitOrigin-RevId: 29f3ca0ae8841f295058feb207e3a225af9a1487
## What changed
Register the disabled-by-default `async_compaction` feature and accept either a boolean or a `[features.async_compaction]` table. Add `model_async_compact_threshold_percent` with a default of 65 and a valid range of 1–100, validated even when the feature is disabled. Setting a threshold alone does not enable the feature.
Expose the threshold in the runtime configuration and update the JSON schema for both configuration forms.
## Testing
Add configuration loader tests covering boolean and table forms, default and custom thresholds, explicit enablement, and rejection of out-of-range values.
GitOrigin-RevId: cf33de76878e7044bc8cf5d146375a07ea9eb078
## What changed
- Reuse `custom_terminal::ModifierDiff` when writing history spans.
- Share the terminal default-color probe and cache implementation between Unix and Windows.
Both changes remove duplicate implementations while preserving existing behavior.
GitOrigin-RevId: 35f8d77ade972b791c6beac0678d2fe2be3aa763
## Why
Policy evaluation and upstream parsers must not select different `Host` field lines from the same request.
## What changed
Return `400 Bad Request` for multiple `Host` headers, including identical values, before policy evaluation, upstream forwarding, or Unix socket routing. Apply the check to both plain HTTP requests and `CONNECT` requests.
## Testing
Add regression tests covering identical and conflicting headers in origin-form, absolute-form, and `CONNECT` requests, verifying rejection without an upstream connection. Also cover single-header forwarding and rejection before Unix socket routing.
GitOrigin-RevId: 8ec95171e120f784629d3121b9a1bedbaf88e7cd
## Why
Saving an Effects preference refreshes the picker. When users toggle repeatedly before saves finish, that refresh can discard newer choices and cause subsequent toggles to use stale values.
## What changed
- Update Effects checkboxes immediately and track pending saves per row, preserving newer choices when an earlier save takes effect.
- Preserve pending Composer stars choices while Animations disables the row, and clear pending state on a regular panel refresh.
- Use the shared list toggle handling for Effects, including Space support, while giving configured key bindings precedence over Space toggling.
## Testing
Add regression tests for repeated toggles across save refreshes, pending Composer stars choices while disabled, clearing pending state, and a Space cancel binding taking precedence over toggling. Update the Effects picker snapshot.
GitOrigin-RevId: 847dcc2dfd40448c7adc56c8c40879db28b1fc39
rka-oai