## Why
Handshake probes can lose upgrade evidence when a subsequent read fails, and diagnostic output can expose URLs, server error messages, and close reasons containing sensitive values.
## What changed
- Retain successful upgrade status and allowlisted tracking headers, including on post-upgrade read failures.
- Return structured probe errors that distinguish typed TLS failures from failures with an unknown stage while preserving the underlying API error.
- Recognize empty close frames as immediate closes with code `1005`.
- Redact wire values from probe debug output and doctor error details, and remove URLs and raw errors from connection logs.
## Testing
Add local TLS and WebSocket fixtures covering certificate rejection, ordinary I/O failures, successful and rejected upgrades, immediate closes, malformed frames, header retention, and redaction. Add doctor output snapshots and error-detail redaction assertions.
GitOrigin-RevId: 1ee53720a7a752043943170d8e4da205ec4e55f3
## What changed
- Build and bundle the standalone executor in `bin/` for both package variants, and advertise its path through `execServer` in `codex-package.json`.
- Include the executor in Cargo and Bazel release packaging, macOS signing and notarization, and Windows signing verification. Preserve prebuilt signed binaries during packaging.
- Make installers detect a missing advertised executor while continuing to accept older packages without it.
## Testing
Add coverage for executor archive contents and executable permissions, signed binary preservation, and macOS entitlement verification. Add package smoke tests for initialization, version reporting, filesystem and process sandbox enforcement, and startup error parity with `codex exec-server`.
GitOrigin-RevId: 5c0027fc19f16cf9c2a5bb52dbd065a9776bc2af
## What changed
Extract executor startup into `codex-exec-server-cli`, shared by `codex exec-server` and the new standalone binary. Preserve executor arguments, authentication, configuration overrides, and sandbox helper dispatch across both entry points.
Build the standalone binary with `cargo build -p codex-exec-server-cli --bin codex-exec-server` and run it with `codex-exec-server --listen stdio`.
## Testing
Add standalone CLI parsing and help coverage, plus an integration test for version reporting, stdio initialization, filesystem reads, process execution, and sandbox helper re-execution when supported.
GitOrigin-RevId: ab3a81c6b91b6f45e4d72d8add2e01af8372e97f
## What changed
- Add `CodexThread::create_child` to create an idle child from a live tool invocation using a host-reserved thread ID and controller.
- Add `prepare_child` and `ThreadManager::create_prepared_child` to capture inherited settings and fresh or forked history, then start the child on another manager even after the parent shuts down. Reject detached preparation when live bindings cannot be transferred.
- Add opt-in `HostToolInvocationLookup` to track live calls and reject duplicate active call IDs within a turn. Revoke invocation access when dispatch ends and clean up provisional children when startup fails or is cancelled.
## Testing
Add tests for invocation lifetime and call ID reservation, inherited policy and live bindings, detached startup with fresh and full history, and cancellation cleanup.
GitOrigin-RevId: 71f17ecb4cd7636517e675df59289a8e8c7105b8
## Why
Image files can become unavailable between turns. Model requests should reflect their current availability while preserving the original references in saved history.
## What changed
- Add opt-in availability checks through `AttachmentStore::is_resolvable`, retaining references by default and on backend or access errors.
- Check distinct file IDs concurrently for each prepared request and replace confirmed-missing images with `[image unavailable]` in messages and tool outputs.
- Apply the same handling to compaction requests, with background image checks running outside the session lock without delaying foreground turn completion.
## Testing
Add unit tests and multi-turn scenarios covering placeholders, preserved history and annotations, concurrent deduplicated checks, backend errors, and nonblocking background compaction.
GitOrigin-RevId: ec98c7014c6e6b5063d87b943d2cd66a9ce7c1e3
## What changed
Add `OutputBoundary` and token-budget formatting APIs that map item indices and UTF-8 text offsets to their positions after truncation. Normalize input positions, remap boundaries within omitted text, and keep end boundaries before generated omission notices.
Unify code mode output formatting: merge text-only output for truncation with a warning, and budget mixed text and audio in input order while preserving images and encrypted content. Accept owned output items to reuse allocations and move retained media without cloning.
## Testing
Add coverage for boundary normalization and remapping, UTF-8 offsets, zero and nonzero budgets, owned and borrowed output equivalence, and allocation reuse.
GitOrigin-RevId: 598407caa88b5a1f84130c4bc2273b85ab02f7f2
## What changed
Add optional `request_permissions` and `auto_review_suffix` fields to catalog approval messages. Use them to customize the permission tool heading and introduce automatic review guidance while preserving the bundled instruction bodies and existing policy gates. Accept `request_permissions_tool` as a deserialization alias.
Limit headings to 500 UTF-8 bytes on a single line and fall back to bundled text for missing or invalid values. Complete catalog approval messages retain precedence.
## Testing
Add coverage for heading validation, serialization aliases, policy gates, fallback behavior, and complete-message precedence. Integration tests verify that model switches append the selected model's headings without rewriting history or repeating unchanged instructions.
GitOrigin-RevId: 3311591a464cc086783fe25271a352dd5a229790
## What changed
Construct the direct transport explicitly in `HttpClient::new` and reuse `without_request_logging` in `new_without_request_logging`.
Bind awaited request results before matching them in transport execution and the Windows proxy retry path.
GitOrigin-RevId: 58b2524491db80ff4d7a954868c1e030a1643a9d
## What changed
Add `MaybeApplyPatch::intercepts_command` to identify commands intercepted by exec without reading patch target files. Recognize explicit patch invocations, malformed patch bodies in those invocations, and implicit raw patches. Share implicit-patch detection with verified parsing, preserving its rejection of implicit invocations.
## Testing
Add a unit test covering explicit, implicit, and malformed patches, plus an unrelated `echo apply_patch` command, without a filesystem.
GitOrigin-RevId: 0bc9f63ca049094a2b17689f4f5585798792fe02
## What changed
- Add `ConnectorRuntimeManager::without_shared_snapshot()` to keep connector catalogs local to each client and disable disk caching.
- Use the thread's own Apps discovery when snapshot sharing is disabled, preserving its catalog across turns.
- Expose `apps_tools_from_mcp_client()` to read a ready, unshared Apps client's tools without starting or waiting for discovery.
## Testing
Extend coverage for shared-cache eligibility and verify that an unshared thread waits for its own Apps discovery and uses its own tools on successive model requests, even when a peer finishes discovery first.
GitOrigin-RevId: d57ddacaede047e67c09d378c86c1aa3577ca56f
## Why
Slow consumers can lose process output or block control replies when the in-process event queue fills. Descendants can also keep output streams open after the child exits, leaving clients unable to tell whether captured output is complete.
## What changed
- Require delivery of `process/outputDelta` and `process/exited` notifications, while allowing process request replies to proceed when event delivery is blocked.
- Exclude time spent delivering output to slow clients from the post-exit drain budget, and drain queued bytes after the deadline closes a stream.
- Keep draining after an output byte cap is reached so the child can finish writing; abort each stream's reader when its drain deadline expires.
- Add optional `outputDrainTimedOut` to `process/exited` and generated bindings. Report incomplete output separately from the child's exit code, treating a fully captured requested byte cap as complete.
- Deliver the exit notification before moving Windows console cleanup to a blocking task.
## Testing
Add regression coverage for full event queues, stalled consumers issuing process control requests, slow output delivery, inherited pipes, output caps, Windows console descendants, and compatibility with exit notifications that omit `outputDrainTimedOut`.
GitOrigin-RevId: 04cfeea5daa7c2b833cdec3f1fff7efeb4d1a86e
## Why
Memory consolidation inherits the session's service tier. Give background consolidation its own setting, defaulting to Standard.
## What changed
Add `memories.consolidation_service_tier` with a default of `"default"` and apply it to the consolidation agent. Normalize recognized tier aliases, including `"fast"` to `"priority"`, and expose the setting in the configuration schema.
## Testing
Add coverage for configuration parsing and alias normalization, Standard consolidation despite an Ultrafast session tier, and an explicit `"flex"` consolidation override.
GitOrigin-RevId: 58031ad56196152d212531d56e7fe9813f83317e
## What changed
Allow `model_catalog_url` under `[model_providers.openai]` while preserving the built-in provider's identity, authentication, and inference routing through `openai_base_url`. Reject other non-default overrides and require an HTTPS catalog URL on an allowed ChatGPT host, without embedded credentials.
Apply the same catalog override merging when checking managed provider requirements for existing threads, preserving each thread's inference endpoint even if user defaults change.
Document the required Codex catalog format, credential forwarding, and existing discovery and access restrictions.
## Testing
Add coverage for allowed and rejected catalog destinations, preservation of built-in provider settings, and managed requirement changes. Add integration coverage for API key and ChatGPT authentication that verifies separate catalog and inference routing.
GitOrigin-RevId: 4d91ada518ebd2a09cbf6ded0ddf4b11c92b77dd
## What changed
- Retain mention targets, image paths, and collapsed paste contents when copying or deleting into the register. Pasting creates fresh element identities and image/paste labels.
- Preserve rich register contents across draft clears and composer replacement.
- Capture `Ctrl+Y` register contents for dot-repeat, restore exact mention bindings through undo/redo, and keep pasted rich objects atomic in Replace mode.
## Testing
Add composer tests and snapshots for repeated rich pastes, register handoff, undo/redo, Replace-mode Backspace, and selection replacement. Add a test ensuring register contents are omitted from debug output.
GitOrigin-RevId: d543d21fd542492af49cce01c2a673e6d3d01664
## Why
HTTP response streams retained copies of completed output items even when inference tracing was disabled, keeping their payloads alive until the response completed.
## What changed
Collect completed output items only when inference tracing or WebSocket last-response bookkeeping needs them. Make the last-response sender optional and create its channel only for WebSocket requests.
## Testing
Add a regression test that verifies an untraced HTTP stream releases a completed item's shared payload after delivery, while the response remains open. Update existing tracing and WebSocket bookkeeping tests for the optional sender.
GitOrigin-RevId: 7695534c6d00f3002005324c43c23bf733f01804
## Why
Terminals that do not answer the default-color probe lose the background shading on submitted prompts.
## What changed
Use an ANSI `DarkGray` background as a fallback. Keep the default text foreground when the background's RGB value is unknown, and preserve the red voice marker in terminal history and transcript selections.
## Testing
Add regression coverage for prompt shading without a terminal palette, voice marker colors under both history wrapping policies, and spoken-prompt selection contrast in compact and detailed transcripts.
GitOrigin-RevId: b8217a2ee1d37a2ea4a011faa75427c881202b82
## What changed
Add `CodexThread::replace_dynamic_tools` to replace the tool catalog without starting a turn. Updates apply to subsequent model steps and future turns while preserving in-flight calls; an empty list clears the catalog.
Require host opt-in through `allow_persistent_dynamic_tool_updates` for persistent threads. Persist replacements before publishing them, restore the latest catalog on resume, and preserve replacements and explicit clears through compaction. Keep explicit resume overrides separate from the persisted catalog.
## Testing
Add coverage for pending calls during replacement, replacement and clearing across restart, compaction recovery, resume overrides, submission ordering after caller cancellation, host opt-in, and persistence failures.
GitOrigin-RevId: fd2634906895a9b72ae7892ddd9435dced45eccb
## What changed
Expose the backend-provided `generationId` on image-generation items and preserve it through legacy completion events, thread history reconstruction, and rollout migration. Default missing IDs to `None` for older history.
Update the generated JSON schemas, TypeScript types, and Python models to include the nullable field.
## Testing
Extend coverage for ID serialization round trips, completed-item notifications, persisted and resumed history, and legacy rollout migration.
GitOrigin-RevId: a7f06faa7ac6833417b257a535d2901c03b79135
## What changed
Expose `literature_search_query` in standalone `web.run` when the active model lists it in `experimental_supported_tools`. Support query text, domain filters, and recency filters, and reject literature search calls for models without the capability.
Pass the active model information to extensions for each step and report literature queries as search actions in web search events.
## Testing
Add coverage for capability rejection, model-dependent schema exposure, app-server round trips with a turn-level model override, and a scenario that searches for a paper and opens the returned reference.
GitOrigin-RevId: 87ef1fadd80ca7bfc2a7642c4e443f9b9f2e0e00
## What changed
Add optional `TurnInputRequest.user_item_uuid` and carry it through turn starts and steering to the recorded user message. When supplied, use `msg_input_<uuid>` as the Responses item ID; inputs without a UUID retain generated IDs. Raw response items ignore this field.
Add `ResponseItemId` helpers to construct and recognize the reserved `msg_input_` namespace.
## Testing
Add regression coverage for steered inputs with identical text and client IDs, checking that each supplied UUID survives recording and UI item IDs remain distinct. Add a scenario test verifying that the assigned ID reaches Responses and persists in subsequent request history alongside an identical ordinary prompt.
GitOrigin-RevId: 440e306bfdaa7facb69b21930c5244a83db2fdae
## Why
A concurrent fork can inherit a PID reservation probe's file descriptor and keep its lock held after the probe closes it, making a stale reservation appear active.
## What changed
Add a drop guard that explicitly unlocks empty PID reservation probes on Unix, including when inspection fails or is cancelled.
## Testing
Extend the duplicated-handle regression test to cover empty PID files, read errors, and cancellation alongside missing PID files. Verify that cancellation occurs while the lock is held and that subsequent status checks do not report an active reservation.
GitOrigin-RevId: 793569c9bcaca02900662a66c33aa48f2bebfd5e
## What changed
Add `ReviewerPool::review_isolated` to run a review in a temporary session without reading or changing the reusable reviewer. Build the request context without a previous session and spawn without a cached snapshot.
## Testing
Add a test verifying that an isolated review receives no snapshot, uses a separate session, and preserves the prewarmed session for the next regular review.
GitOrigin-RevId: 6cceff0c399f30f3c3303b0bf5232c274c902e3a
## What changed
Keep original scheduled messages and heartbeat provenance during synchronous transcript collection. Add `prepare_sync_context` to compress repeated heartbeat instructions across the complete collected transcript before selecting the requested suffix, preserving reference numbering in full and delta prompts.
Route synchronous reviewer prompts through this preparation step and reject composition of unprepared scheduled messages. Move transcript profiles and REPL evidence rendering into the preparation module. Asynchronous heartbeat compression remains in collection.
## Testing
Extend heartbeat transcript coverage to compare prepared full and delta prompts with the existing rendering and verify that direct composition rejects unprepared scheduled messages.
GitOrigin-RevId: c9824b8862021301a1f537a74f83bc7eb20aa0ad
## What changed
Add selectable `profile` and `provider` items to the TUI status line, including setup previews and theme colors. Render them as `profile: <name>` and `provider: <id>`, omitting the profile when unset and the provider until a thread is active. Accept `model-provider` as an alias for `provider`.
## Testing
Extend status line tests to cover omitted values, the active thread's provider, provider updates, and a selected profile. Update the setup popup snapshot to include both items.
GitOrigin-RevId: e7cbb7fb1e7017cfbd6ee71e1df7d0ddb1836ed4
## What changed
Use directory-entry file types when expanding volume-root grants, and resolve metadata only for symbolic links. Skip links whose targets cannot be resolved, while preserving grants for ordinary children and valid links.
## Testing
Extend the volume-expansion test to cover dangling and self-referential links alongside an ordinary file, a valid symbolic link, and an explicit deny grant.
GitOrigin-RevId: 8acb4072a1f5edd4351738b5780ac659fdf94f01
## What changed
- Add `accountSession/list` and `accountSession/switch` for the active managed ChatGPT login, with optional workspace membership refresh.
- Validate membership and workspace policy, pause new turns, and unload running threads before switching credentials. Notify clients with `thread/closed` and `account/updated`; selecting the current workspace leaves threads running.
- Support browser sign-in for SSO through `allowBrowserLogin`, binding the callback to the requesting login and workspace. Require a separate sign-in for government workspaces.
- Preserve the selected workspace across token refresh and conditionally restore the previous workspace if routing validation fails. Guard credential writes against concurrent login or logout changes.
## Testing
Add app-server and auth tests covering workspace switching, credential preservation, policy and identity rejection, conditional rollback, stale sessions, and SSO callback validation.
GitOrigin-RevId: 1234abc33b71e69b66318da3e201fccc08867906
## Why
Replacing the Node process lets the native binary receive signals and report its termination directly.
## What changed
- Use `process.execve` on non-Windows platforms when available, preserving CLI arguments and the prepared environment.
- Install a `SIGXFSZ` handler before replacement so the native binary gets the default signal disposition instead of inheriting Node's ignored signal.
- Retain asynchronous spawning for Windows and older Node versions.
## Testing
Extend the exit-status test to check that `execve` leaves no child process and that `SIGHUP` and `SIGXFSZ` terminate it with the original signal, while retaining coverage for the spawn fallback.
GitOrigin-RevId: 7437f15af419fce5b10c1c00425c7d38160991fc
## Why
A concurrent fork can inherit a PID reservation probe's file descriptor and keep its lock alive after the probe closes it. This can make the app-server daemon appear to be starting even when no live startup reservation exists.
## What changed
Explicitly release the probe's lock on Unix before returning that no reservation is active, and propagate unlock failures.
## Testing
Add a regression test using a duplicated file handle to simulate fork inheritance and verify that the probe leaves the daemon inactive. Extend coverage to confirm that a held reservation still reports `Starting` when the PID file is missing, and include pruning results in release-retirement assertion failures.
GitOrigin-RevId: ecc58df1fe5cadecb6ed658352f1300c39237358
## What changed
Move planned-action construction from the core Guardian prompt builder into the exported `codex_guardian_reviewer::build_guardian_planned_action` helper. Keep the existing action kinds, MCP tool and connector descriptions, and reason selection unchanged.
GitOrigin-RevId: 119323523ed2e199c998c1139fb8e6255bc730d9
## What changed
- Add `--codex-ui-plugin-process` to `codex-code-mode-host` and a typed protocol for initializing, invoking, and closing UI plugin isolates.
- Support persistent module state and string results from synchronous or promise-based exports across up to 32 isolates in one helper process.
- Require sandbox-enabled V8 with JIT disabled, restrict executable loading and globals, and enforce execution deadlines and per-instance queue, source, result, and promise limits.
- Close affected isolates on script failures, cancellation, or overload, and return bounded script diagnostics.
- Bump the helper handshake to version 3 and increase the maximum frame size to 1 MiB.
## Testing
Add wire-level integration tests and runtime unit tests covering promise settlement, cancellation, loading restrictions, diagnostics, failure isolation, capacity reuse, transport bounds, and watchdog synchronization. Add CLI tests for argument validation before helper startup.
GitOrigin-RevId: bad6a57c84d9197c1fd76a78cf060502c5bc8bb8
## What changed
Move executor path-convention selection from the approval caller into `ReviewAction::from_approval_action`, using the supplied `GuardianReviewContext`. Preserve Windows and POSIX rendering based on the executor OS, with path inference for legacy executors that do not report an OS.
## Testing
Update the command approval test to exercise `ReviewAction` with Windows, Linux, and legacy executor contexts, checking both approval JSON and assessment working directories.
GitOrigin-RevId: b13fad1d4033976cb94b899b62a3779c97e171fa
## What changed
Move approval-request assessment and analytics mappings, target item lookup, and turn ID selection into `codex-guardian-reviewer`. Export the helpers from the crate and re-export them in core, preserving existing behavior.
## Testing
Move the apply-patch serialization test into the reviewer crate. It checks that Windows paths are preserved and patch text is omitted from the assessment action.
GitOrigin-RevId: 3a011db0bc51c653bb0877c4b0cb52eb61c63244
## Why
Copying in a WSL SSH session with X11 forwarding should use the forwarded display's clipboard instead of writing to the Windows host clipboard.
## What changed
Treat a nonempty `DISPLAY` in an SSH session as X11 forwarding and bypass the Windows clipboard backend on WSL. Keep terminal clipboard forwarding in place.
## Testing
Add a regression test that verifies a WSL SSH session with X11 forwarding uses the native clipboard without invoking the Windows clipboard backend.
GitOrigin-RevId: 29f98dc39e5851d63d0459a55e234346167da26f
## Why
macOS `lsof +D` can return status 1 even when it finds open files, causing release cleanup to be skipped. It can also escape non-ASCII installation paths, preventing full-path matching from recognizing releases in use.
## What changed
Accept `lsof` status 1 while preserving matches, and continue to reject other failures or any stderr. Match the installer-owned ASCII path suffix on macOS and recognize processes whose working directory is the release root. Apply these checks before and after moving a release for deletion.
## Testing
Add coverage for `lsof` exit statuses, output, and errors. Exercise release retention with a non-ASCII installation path and, on macOS, a process holding the release directory open.
GitOrigin-RevId: 06181467c7d1bb256b6781241a444976b890d7c9
## Why
Restoring killed text should restore its local image attachments along with their placeholders.
## What changed
Keep image paths in the composer's kill buffer, including across composer replacement, and restore attachments with fresh labels when text is pasted back. Preserve plain-text insertion in Vim Replace mode so per-character recovery continues to work.
## Testing
Add regression coverage for restoring images after submission, transferring the kill buffer between composers, and preserving existing images and Backspace recovery in Vim Replace mode.
GitOrigin-RevId: 9bc052a1b372d5a1245ddc54c7ffafee4a637494
## What changed
Recognize `\%` as a literal `%` in inline and display math, so expressions such as `$3.8\%$` render as `3.8%`.
## Testing
Extend inline and display math snapshots to cover escaped percent signs, including an inline expression with multiplication and a fraction.
GitOrigin-RevId: a0f9935dc8492fe1cd5a0d33d62930c48c5433b8
## What changed
Add the disabled-by-default `omit_skill_descriptions` feature flag to omit descriptions from model-visible host, executor, and cloud skill catalogs while retaining names and locators. Adapt catalog introductions, usage instructions, and alias budget accounting to the selected policy.
Persist the description policy in catalog snapshots and refresh usage instructions when it changes. Ignore description-only edits in catalog deltas while descriptions are omitted.
## Testing
Add coverage for alias token budgets, policy changes with older snapshots, and suppressed description-only updates. Extend scenario snapshots to cover plugin visibility and catalog deduplication with descriptions omitted.
GitOrigin-RevId: 5bf1dcee8c04059c7a78998affc1efcdef8ec883
## What changed
- Retain captured root conversation sources until synchronous preparation applies request limits. Keep asynchronous snapshots limited to the bounded messages.
- Move approval-reason truncation and tool-description escaping and truncation from composition into preparation.
- Extract synchronous image preparation and admission into `guardian-context`, preserving the prompt token cap and model context-window checks.
GitOrigin-RevId: d3a864c77d8f16bd3e77df436d80c4d2738c43cc
## What changed
Move approval request serialization and model-visible action formatting from `codex-core` into `codex-guardian-reviewer`, expose both helpers publicly, and re-export them from core. Preserve the existing JSON shapes and formatting.
## Testing
Move the MCP tool-call and network-access trigger serialization tests into the reviewer crate, retaining coverage for omitted optional fields and Windows path rendering.
GitOrigin-RevId: c0093c023b2f4c1be16bac481f1173a5da6ac25c
## What changed
Export `GuardianPromptItems` and `compose_guardian_prompt` from `codex-guardian-reviewer`, and delegate prompt composition from core to the new helper. Core continues to collect context sections and choose the history version.
Preserve full and delta transcript selection, empty-transcript placeholders, cursor accounting, and Node REPL evidence sequencing.
GitOrigin-RevId: e3e2b2a87a42709ddb5441f47cbb7461b0130579
## Why
Migrating selected threads reads rollout headers even for unrelated threads. Use the state database to skip files whose indexed owner and canonical filename agree that they belong to an unselected thread.
## What changed
- Read thread IDs and rollout paths without creating or migrating the state database, and fall back to header inspection if the index cannot be read immediately.
- Keep inspecting selected, unindexed, misnamed, and conflicting paths so rollout headers still determine ownership.
- Recognize stable thread IDs in reverted rollout filenames and match plain and compressed paths when filtering.
## Testing
Add a regression test covering pruning of unrelated indexed rollouts, including archived and compressed files, while retaining header inspection for selected and ambiguously named files.
GitOrigin-RevId: d76827174c00f5bc9e38ad9b69dadeee91f4415f
## What changed
Export `GuardianApprovalRequest`, `GuardianNetworkAccessTrigger`, and
`GuardianMcpAnnotations` from `codex-guardian-reviewer` and re-export them in
core. Move environment lookup and scope classification with the request type,
preserving existing classification and credential-safe debug output.
GitOrigin-RevId: 5253a5124d5e370a1230b46aa7dc48b24b1c902e
## What changed
Move Guardian session-source detection, the reviewer name constant, and parent-thread prompt cache key generation into `codex-guardian-reviewer`. Re-export them from core, preserving existing behavior.
## Testing
Move the existing prompt cache key test into the reviewer crate, retaining coverage for parent-thread scoping, key length, and missing or inapplicable inputs.
GitOrigin-RevId: c14b3c030ba93676a0799aba56b23a676dbbc56d
## What changed
Extend the incremental Guardian review scenario with historical encrypted `send_message` calls in the default and custom tool namespaces. Add snapshots and assertions that both reviews retain each call and parent reply exactly once, in order, preserving call IDs, namespaces, and encryption metadata with a paired evidence-only output.
Verify that successive reviews reuse the same thread and append to the previous input. The scenarios use mocked ciphertext and decisions to check transport behavior, not backend decryption or authorization outcomes.
GitOrigin-RevId: a0927531b6cdae496fefa2a7f35032f7413807b6
## What changed
Add mock guardian approvals for setup handoffs in `handoff_scenario` when native multi-agent history is enabled. Match only the trailing `function_call_output` so historical handoffs cannot consume the final command's separate denial response.
Require five approval requests with native history enabled and none with it disabled.
GitOrigin-RevId: 7b14c97aa85adc63c4d3bf95014c44831eae1c85
## What changed
Make tool descriptions precede namespace descriptions in Code Mode execution and tool search. Mark `code_mode_tool_description_first` as removed and simplify cached definitions to depend only on the input schema budget.
## Testing
Update Code Mode integration assertions and catalog snapshots to expect tool descriptions before namespace descriptions.
GitOrigin-RevId: 1344d4c264d1d2b18ee0388a064b6813ca81f770
## What changed
Export `token_usage_delta` from `codex-guardian-reviewer` and use it in the core review session. Preserve the existing calculation, including clamping negative deltas to zero.
## Testing
Move the existing nonnegative token usage delta test alongside the helper in the reviewer crate.
GitOrigin-RevId: fcfce24b8d50a63dcd1122a4204939bc6750feda
## What changed
Add `GuardianScope::for_multi_agent_tool` and use it for both tool review requirements and asynchronous scoring. Classify `spawn_agent`, `send_message`, `followup_task`, and `interrupt_agent` as native tools when they match the configured namespace, accepting default namespace spellings when none is configured. Keep review enablement checks at each call site.
GitOrigin-RevId: 9180ccdc629532f6c171d899dc5881e15b447141
## What changed
- Recognize `policy_capacity_unavailable` in HTTP and streaming API errors and expose it as `policyCapacityUnavailable` through the app-server protocol and generated bindings.
- Display “Policy-restricted capacity is unavailable. Please try again later.” while preserving overload retry behavior based on `Retry-After`.
- Handle the new error in TUI overload rendering, reviewer retries, and error telemetry.
## Testing
Add coverage for HTTP and streaming error classification, retry advice, WebSocket terminal error attribution, protocol serialization, and TUI rendering with and without an upstream message.
GitOrigin-RevId: 1b9536fba1e11cc891dcf689aee35410d82f5412
dob-oai