## Why
Search discovery must account for the active model's supported tools while honoring host search restrictions. Automatic fallback must also preserve the provider that owns existing browsing references.
## What changed
- Resolve search from model metadata, client configuration, and managed requirements, exposing the selected provider's tools.
- Discover hosted search with the model's supported tools and carry the discovered invocation context into calls.
- Allow native fallback for eligible discovery failures only when search policy permits it and hosted search has not already been used. Do not replay failed hosted invocations through native search.
- Keep native fallback references on native search across later turns, discovery recovery, and resume. Release unused fallback offers and coordinate background calls to prevent conflicting provider use.
## Testing
Add app-server and core scenario coverage for provider selection, discovery failures, search restrictions, reference continuity, resume, and concurrent background search. Include a snapshot covering discovery and call history across a model switch.
GitOrigin-RevId: 0f22605722daaa424c4fa34660654cde5a6db049
## What changed
Once Responses requests fall back to HTTP after exhausting WebSocket retries or receiving HTTP 426, keep subsequent requests and turns on HTTP for the remainder of the session. Clear cached WebSocket state and remove background recovery probes, recovery trials, and their acknowledgement timeout.
## Testing
Update fallback tests to verify that a second turn stays on HTTP without additional WebSocket attempts. Adjust retry and `Retry-After` request-count assertions to exclude background probes.
GitOrigin-RevId: a4f6376604527c14db8d60f79bd21c3aaf289f70
## Why
Injected agent messages need to reach active sampling before stale tool calls execute, while messages queued around a final answer must wait for the next task without requesting another response.
## What changed
- Include injected `agent_message` items in mailbox preemption checks when the matching turn still accepts mail and `DeferMailboxPreemption` is disabled.
- Treat injected agent messages as queue-only mail so they do not reopen delivery after a final answer or force a follow-up response.
## Testing
Add a streaming scenario and context snapshot covering idle injection, active sampling preemption, deferred delivery around a final answer, and preservation of message payloads and ordering before the next task.
GitOrigin-RevId: 70f49cbe5074725c11e1fcb33eea05e0e92e17f7
## Why
Host-owned runtimes skip controller rollout-budget accounting. Give hosts an optional callback to account for inference spending and signal budget exhaustion.
## What changed
- Add `AgentUsageAccounting`, `AgentUsageReport`, and `UsageAccountingDecision`, captured at startup and supported by host child spawn reservations.
- Await per-response usage reports with thread and turn attribution, including provider budget units and foreground and background compaction usage.
- Map host exhaustion to `SessionBudgetExceeded` and accounting failures to fatal errors without retrying inference. Preserve the original error for incomplete responses.
- Keep controller accounting unchanged and leave accounting disabled for host-owned runtimes without a provider.
## Testing
Add coverage for usage attribution, child accounting, budget exhaustion, nonretryable accounting failures, incomplete responses, and compaction reporting, including background usage charged once to its originating turn.
GitOrigin-RevId: 887703f50fe882497e4e8cb33259072935ec82e0
## What changed
Add `AgentRuntimeOwnership::Host` and an `AgentTurnAdmission` provider so hosts can acquire execution capacity before runtime tasks run, including root turns. Require the provider at startup and retain its execution guard for the task lifetime. Controller ownership remains the default; host ownership skips controller rollout-budget accounting, budget reminders, and subagent roster context.
Keep pending admission interruptible without blocking input submission. Report admission failures through the turn lifecycle, record accepted input before exposing the error, and unwind task setup such as review mode. Preserve recoverable input during suspension. On interruption, skip prompts that still require a `UserPromptSubmit` hook rather than recording them without inspection.
## Testing
Add integration tests for missing-provider startup cleanup, denial and immediate retry, steering during admission, cancellation, execution guard lifetime, submit-hook handling, suspension recovery, and review-mode cleanup. Add a context snapshot verifying retained input reaches the next admitted turn.
GitOrigin-RevId: 9ee85fad2ecfbfb38735bac8ef88888fb179a967
## Why
A transient WebSocket failure previously kept the session on HTTP for its remaining lifetime, even if WebSocket connectivity recovered.
## What changed
- Probe WebSocket connectivity in the background during HTTP fallback, allowing a subsequent request to use a ready connection without delaying HTTP requests.
- Verify recovery with a real request and bound its first-response wait to five seconds. Keep HTTP 426 responses and recovery timeouts on HTTP for the rest of the session.
- Keep retries of an HTTP request on HTTP. If a recovered connection fails before output, replay the request over HTTP without consuming its ordinary retry budget; failures after output use normal retry accounting.
- Rate-limit recovery probes and honor `Retry-After` deadlines.
## Testing
Extend integration coverage for successful recovery within a turn, disconnects before and after output, full-context replay followed by incremental WebSocket requests, persistent HTTP 426 fallback, and retry deadlines.
GitOrigin-RevId: 1c139a4bbce1a67442db5622fd70ff0cc8056c9c
## What changed
- Add `model_messages.plugins` overrides for generic usage guidance, selected-plugin hints, and recommendation introductions, with independent fallback to bundled wording for missing or invalid fields.
- Validate text limits and required template placeholders while keeping plugin identities, available capabilities, and eligibility controlled by the client.
- Track generic guidance changes across model switches and session resume, emitting replacement or retirement notices and supporting legacy snapshots and retained history.
## Testing
Add unit and scenario coverage for sparse overrides, literal text preservation, template validation, UTF-8 size bounds, capability filtering, recommendation fallback, and guidance recovery after resume.
GitOrigin-RevId: 148b658c2d8490c9d0dc208addbee1be5342a5f5
## What changed
- Allow audio inputs to carry uploaded file references via `fileId` in the app-server API and `file_id` in model content, preserving existing inline audio shapes and Python class names.
- Preserve mixed inline and file-backed audio ordering through history reconstruction and migration, and budget file-backed audio conservatively when duration is unknown.
- Keep uploaded audio disabled at runtime: reject it before inference or raw history injection, preserve accompanying text, and prevent rejected attachments from reaching follow-up or resumed requests.
## Testing
Add coverage for wire-format round trips, SDK compatibility, mixed audio history ordering, and rejection during turn start, steering, task completion, and raw injection.
GitOrigin-RevId: 70895c47a2494c0c9936348ad1250eb1b7b892a6
## Why
Retained output from failed responses or responses without usage data can be omitted from context token estimates, delaying automatic compaction even as reasoning accumulates.
## What changed
Track the history boundary covered by the last usable usage report and estimate all items added after it. Spending-only updates preserve that boundary, while successful usage reports cover the request input even when the response contains no output items. Avoid counting unreported reasoning twice and align recomputed context counts and remote compaction estimates with the boundary.
## Testing
Add unit coverage for initial input, empty responses, spending-only updates, and reasoning accounting. Add regression coverage verifying that accumulated unreported reasoning triggers compaction for incomplete responses with or without usage and completed responses without usage.
GitOrigin-RevId: 80c0b800c7236ff40ae53fdd18320e0d06e026ab
## Why
A network request can disappear while its approval prompt is still pending. Leaving that prompt active allows a late cancellation to interrupt the turn even though the request no longer needs approval.
## What changed
- Track network prompt lifetimes in core and notify the app server when a prompt is resolved or withdrawn.
- Cancel the corresponding server request, emit `serverRequest/resolved`, and release its permission guard. Resolve outstanding network prompts during thread shutdown as well.
- Claim a live pending approval and validate its originating turn before applying a decision, so stale responses cannot interrupt a turn or persist policy changes.
- Distinguish network approvals that require the originating turn from proxy approvals that can be serviced by a later active turn.
## Testing
Extend regression coverage for disconnected network requests, late cancellation responses, turn completion during review, shutdown cleanup, mismatched turn IDs, and command approvals from background code running during a later turn.
GitOrigin-RevId: 78f656d92607ec21ddd0c5d781ed0947588cb2eb
## Why
Cyber access program selection was restricted to the `openai` provider ID, preventing custom providers from accepting explicit program selections.
## What changed
- Add `capabilities.cyber_access_programs` to provider configuration and preserve it through remote thread configuration. Default to enabled for `openai` and disabled for other provider IDs, with explicit overrides taking precedence.
- Use the capability when forwarding `access_programs.cyber` and determining Daybreak availability in exec and the TUI. Refresh TUI eligibility when the server changes the selected provider.
- Preserve authentication, API-key feature gates, and catalog access checks; the capability does not grant program access.
## Testing
Add coverage for capability defaults and overrides, configuration round trips, custom-provider requests with both authentication types, API-key feature gating, compaction, WebSocket response reuse, exec selection, and TUI provider changes.
GitOrigin-RevId: 579f7c33c9e857ff5fedd08311aa8e3129740c1f
## What changed
- Match final serving assessments to their response and output item, then retain them with tool invocations and code cells across waits.
- Use matching assessments instead of asynchronous scoring for adaptive review: low-risk actions can skip synchronous review, while elevated-risk actions require it.
- Revalidate captured authorization and policy before approval. Require fresh review for stale or incomplete authorization, preserve explicit review requirements, and discard assessments when hooks rewrite tool input.
- Keep asynchronous scoring as the fallback when an assessment is missing or does not match.
## Testing
Add integration coverage for direct and nested tool calls, hook rewrites, explicit escalation, interrupted responses, catalog-driven policy, incomplete authorization, and assessment retention across code-cell yields and waits.
GitOrigin-RevId: 4346a227204137f018c03c34b302d6f1e9592c96
## What changed
- Add the disabled-by-default `interrupt_on_cancel` feature to interrupt upstream inference when a Responses Lite turn is canceled, with a one-second cleanup deadline.
- Detach canceled WebSocket responses so replacement requests can open a new connection without waiting for cleanup.
- Add an optional `interrupt_on_cancel` override for thread samples, inheriting the parent setting when omitted.
## Testing
Add coverage for cancellation during inference and steering, replacement turns proceeding before cleanup finishes, and prompt cancellation of model response interceptors.
GitOrigin-RevId: 501368746fc04fd3fdbd45c1f34c75ce57f4e955
## Why
Resuming or forking a thread loses the session flag indicating that reported token usage already includes retained reasoning. Counting that reasoning again can trigger automatic compaction too early.
## What changed
Persist `last_usage_includes_reasoning` with `TokenUsageInfo` and use it when estimating context usage. Record the flag with completed sampling and compaction usage, and mark local estimates as including reasoning. Older records without the field retain the existing accounting behavior.
## Testing
Add coverage for saved inclusive, exclusive, and legacy token counts. Extend the encrypted-reasoning compaction test to verify the compaction threshold for live, forked, and cold-resumed threads.
GitOrigin-RevId: 9f22262a693111ec0406dce11804eb4844ad8983
## What changed
Decode `response.metadata` Guardian assessments into typed `GuardianAssessment` events over SSE and WebSocket, carrying `response_id`, `item_id`, and `requires_review`. Ignore assessments with missing, empty, or incorrectly typed required fields while preserving independent delivery of other metadata.
Expose the assessment type to request-scoped extension interceptors. Keep assessments out of conversation history; they do not authorize tool execution.
## Testing
Add SSE coverage for assessment ordering, malformed metadata, and mixed metadata delivery. Extend model request tests to cover SSE and WebSocket interceptor delivery and exclusion from subsequent model input.
GitOrigin-RevId: 282b8817ee97346d76b40bcb412eef023df60ff1
## Why
Shared routing cookies can change after a parent connection is established. Guardian reviews need the routing affinity captured with their parent response so unrelated traffic does not replace that hint.
## What changed
- Capture `__oailb` and `__cflb` cookies from successful ChatGPT WebSocket handshakes and pair them with the parent response ID.
- Retain the snapshot throughout a review and its retries, and apply it only to Guardian requests with the matching parent response. Reconnect cached reviewer sockets when the parent affinity changes.
- Preserve cookie expiry and restrict reuse to the original secure origin and endpoint path. Replace routing cookies while retaining unrelated cookies, and keep the captured affinity out of prompts, history, and serialized metadata.
## Testing
Add coverage for cookie scope, expiry, duplicate rejection, and merging; replaying a captured pin after shared cookies change or are deleted; and retaining the matching parent snapshot through compaction without exposing cookies in review prompts or metadata.
GitOrigin-RevId: d180679e6265b612a3617f20ca1f5dec89e73aa8
## Why
Legacy notifications only need user text, but retaining the full sampling input keeps image and audio payloads alive while the response streams.
## What changed
Extract notification text from the first request attempt only when `notify` is configured, and retain those strings across retries. Consume and drop the prompt after setting up the response stream, excluding attachments and media wrapper labels from notification text.
## Testing
Extend the notification integration test to verify that `input-messages` contains only user text while the model request still includes image and audio attachments.
GitOrigin-RevId: af4653c0e7ac2bf7006f5ae5a82a3e724c30c734
## 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 `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
Add the `features.simplified_selection_hints` flag, disabled by default. When enabled, replace detailed plugin capability instructions with a concise developer message identifying the selected plugin and directing the model to use its available skills and tools.
Continue omitting hints for plugins without available capabilities and bounding hint length. Preserve the existing detailed instructions when the flag is disabled.
## Testing
Cover plugin refresh in an existing thread with the flag enabled and disabled. Verify that explicit skill prompts precede the simplified plugin hint and that the hint uses the developer role.
GitOrigin-RevId: 64fc3347bd49fffe675f343a4733ced1b4e5e88c
## What changed
Add the default-disabled `critical_path_timing` feature to capture dependency paths for root-agent turns. Connect direct tool dispatch, execution-lock handoffs, and completed tool batches to turn timing so overlapping work shares its common history without double counting.
Omit captures for unsupported dependencies, including CodeMode, agent communication, approval review, and user waits. Settle aborted tasks before freezing timing, invalidate captures on forced teardown, and apply runtime feature changes only to future turns.
## Testing
Add coverage for serial and overlapping tool dispatch, execution-lock dependency handoffs, cancellation, unsupported paths, frozen timing after completion, and configuration refresh boundaries.
GitOrigin-RevId: 71458b654b1c4a05799886928b449c948ea33109
## What changed
- Record time spent waiting for each stream event in `codex.response.receive_wait_ms` on the existing `handle_responses` span.
- Add trace spans for startup prewarm consumption, turn-start lifecycle callbacks, assistant text flushing, and response stream error handling.
- Include the turn ID and lifecycle phase on the turn-start span and the request type on the error-handling span.
GitOrigin-RevId: 7792508c830ef05b03aed01eb840ca7e6577ad0b
## Why
Prompt telemetry can include full user text. Emitting it before hook acceptance logs prompts that `UserPromptSubmit` hooks subsequently block.
## What changed
Move prompt telemetry from input submission to hook acceptance, including queued and steered input. Preserve telemetry for accepted Guardian review prompts that are converted to response items during finalization.
## Testing
Extend hook regression tests to assert that blocked prompts are absent from telemetry and rollouts, while accepted queued prompts are logged exactly once. Extend Guardian review coverage to assert exactly one prompt event across inspection steps.
GitOrigin-RevId: aa71f98de2356dd08f62ff492e897ab3a43b54c9
## What changed
Add the disabled-by-default `guardian_recorded_items` feature to forward recorded reviewer items and interruptions through the parent's raw-event subscription, preserving reviewer thread and turn IDs without modifying the parent's history or completing its turn.
Emit reviewer items as `rawResponseItem/completed` notifications and interruptions as interrupted `turn/completed` notifications. Mark opted-in reviewer requests with `guardian_recorded_items` metadata.
## Testing
Extend Guardian integration tests to cover reviewer identity, recorded-item metadata replay, concurrent reviewer interruptions, independent parent completion, and behavior without the feature enabled.
GitOrigin-RevId: beb4dbf6c6f2a800a2fbf3cfd11ae174a1382386
## Why
Responses Lite includes retained reasoning in its reported token usage even without an `X-Reasoning-Included` header. Adding a separate reasoning estimate can trigger automatic compaction prematurely. Locally recomputed usage also already includes retained reasoning.
## What changed
Initialize reasoning accounting from `use_responses_lite` before each sampling or compaction request, while allowing server events to override it. Mark locally recomputed token usage as including reasoning.
## Testing
Update the compaction regression test to verify that Responses Lite usage without the header stays below an 81-token limit at 80 tokens and triggers compaction at 81 tokens.
GitOrigin-RevId: 06f9272cb3c259e927941f1d2a1d7096310ffb2c
## Why
Mailbox preemption can discard completed response items and lose the response ID needed for incremental continuation. Buffered tool calls must also wait for queued agent messages to be read.
## What changed
- Gracefully interrupt and drain Responses Lite streams, retaining completed items and the response ID for the next request.
- Cancel buffered tool calls before dispatch and suppress partial items that may be discarded during the drain.
- Keep user steering queued until draining finishes. Include pending mail in follow-up requests after disconnects and preserve content-filter recovery guidance.
## Testing
Add WebSocket regression coverage for buffered completion, interrupts without acknowledgement, buffered tools and partial items, user steering during draining, disconnects, and content-filter recovery.
GitOrigin-RevId: c9d3fb87a81850e5686e07f28103e2ff4378ddec
## Why
Resuming a thread can replace saved model settings with global defaults. A selected service tier can also be lost when the current model does not support it.
## What changed
- Restore saved service tier, reasoning summary, and personality on resume unless the caller explicitly overrides them, respecting thread ownership of settings snapshots.
- Retain the selected service tier for persistence, settings notifications, and resume comparisons while filtering the effective tier against the current model's capabilities.
- Carry the runtime model, service tier, and personality into thread revert reconstruction.
- Persist initial thread settings before user input during MCP startup so an interrupted first turn retains its settings checkpoint.
## Testing
Extend regression coverage for cold resume and explicit overrides, unsupported service tiers, settings preservation during revert, and settings checkpoints before user input when MCP startup is interrupted.
GitOrigin-RevId: e5ec012a5678048c5eedd443e524f38c142bf9fd
## What changed
Add an `incremental_tools` attribute to `codex.tools.registered` using the effective mode for the current sampling request. Compute the mode once and reuse it for registration metrics and prompt construction across transport retries.
## Testing
Extend integration tests to verify the attribute for incremental and non-incremental requests, including when the feature flag is enabled but the Responses API does not support incremental tools. Assert registration sample counts and retain coverage that stream retries do not duplicate registration metrics.
GitOrigin-RevId: e5b7a1075fe722ef6bdf900ae07e2b1b06aecfb5
## What changed
- Build executor skill catalogs from discovery V2 inventories without reopening host paths or falling back to legacy discovery.
- Merge Cloud and executor skills by remote plugin identity and skill name, retaining Cloud metadata and accepted provider read routes. Apply plugin selection and disablement before merging.
- Share a stable catalog across each step's rendering, list/read tools, and explicit and implicit skill invocations.
- Preserve unmatched executor entries when Cloud catalogs are unavailable or partial; only complete Cloud catalogs disable unmatched remote entries.
- Bind discovery caches to captured environments and executor sessions, and refresh MCP when V2 inventories change. Preserve pending invalidations when forced MCP refreshes fail.
## Testing
Add regression coverage for V2 inventory sharing and refresh, environment isolation, catalog merging, partial Cloud results, disabled plugins, and catalog consistency across steps and read routes.
GitOrigin-RevId: 842bdcbc9e19364dd69a9abf9e48330b7425a76c
## What changed
- Add `PreModelRequest` and `PostModelRequest` hooks around model request attempts, including retries. Pass model, session, turn, and optional subagent context; report completion, failure, or cancellation with errors and token usage when available.
- Treat hook output as diagnostic data without allowing it to veto requests or alter model context. Wait for synchronous post-request hooks before dispatching tools, while asynchronous hooks allow streaming tool dispatch to continue.
- Apply memory-consolidation hook filtering and expose the new events through configuration, protocol schemas, generated clients, and the hooks browser.
## Testing
Add integration coverage for failures and retries, ignored control directives, cancellation, synchronous and asynchronous tool ordering, interrupted-turn resume and token accounting, and memory-consolidation filtering. Update hook schema fixtures and hooks browser snapshots.
GitOrigin-RevId: 6ed7b39fc8bb94cf467e350d07d97fcc729c21ce
## Why
Contextual fragments and extension-owned world-state sections currently return strings, limiting them to text content.
## What changed
- Return `ContentItem` from fragment bodies and rendering, preserving classification and attribution. Apply text markers only to `InputText`; pass other content through unchanged.
- Allow extension prompt and world-state contributions to carry typed content, and preserve media in context-window hints.
- Hash non-text world-state fragments by their serialized content and role while retaining the existing text hashing behavior.
- Add shared content-to-tool-output conversion that preserves image references, image detail, and audio.
## Testing
Add coverage for typed world-state hashing and preservation of image content and attribution. Expand code-mode output tests to cover input text, output text, file-backed and inline images, and audio.
GitOrigin-RevId: 2ed00d466a3f169b151202f3eef71e667ea64be1
## 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
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
## 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
## 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
## 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 the disabled-by-default `output_token_replay` feature to request `output.encrypted_content` from OpenAI providers.
- Preserve encrypted message and tool-call output, message and function-call status, and output-text `annotations` and `logprobs` through history and replay. Update the generated protocol schemas and SDK types.
- Strip output ciphertext from requests to other providers while retaining it in stored history, including after replay is disabled. Remove encrypted content from memory extraction inputs.
## Testing
Add integration coverage for the feature/provider combinations, metadata replay across follow-up requests, and history preservation when resuming with replay disabled.
GitOrigin-RevId: 77729a021b6771933402aebe74ece8eb39c057e2
## What changed
Add the `codex.tools.registered` histogram to count finalized tool registrations by `exposure` and `tool_mode`, including zero counts for empty exposure categories. Record it once per sampling request before transport retries, excluding hosted tools. These counts describe the registered inventory, not model visibility or tool reachability.
## Testing
Add an integration test covering a stream retry, direct and deferred namespace tools, a model-only tool, and hosted web search. Verify that registrations are recorded once, hosted tools are excluded, and metric labels contain no tool names or descriptions.
GitOrigin-RevId: 02b450743886e3d487d6c3bcda12eb65f6d1104b
## Why
Published MCP elicitation requests could lose their response waiter without notifying the event consumer that the request was no longer answerable.
## What changed
- Add `ElicitationAbandoned` with the original server name and response token for requests published through unbounded event channels.
- Emit abandonment exactly once when a waiter is dropped, response delivery fails, or the MCP runtime shuts down.
- Close the elicitation router during shutdown, reject new requests on unbounded sources, and wait for terminal deliveries before shutdown completes. Preserve bounded sources' request-only behavior and backpressure.
## Testing
Add router tests for ordered, exactly-once abandonment, closed receivers, failed response delivery, shutdown coordination, and bounded-channel backpressure. Add a core integration scenario verifying that a pending form's abandonment precedes `ShutdownComplete`.
GitOrigin-RevId: 53fe9f4a8b32f7324be180288aa492f31f4d2057
## Why
Skill selection and disabled-path matching need to recognize equivalent Windows paths across case and separator differences while preserving literal spaces, `%`, and `#` in native filenames.
## What changed
- Carry `PathUri` through skill metadata, discovery mappings, filesystem lookups, disabled-path sets, and invocation analytics.
- Match linked skill mentions and TUI selections by parsed path identity.
- Use `LegacyAppPathString` for app-server skill paths and regenerate schemas and SDK types, preserving native path strings on the wire.
- Preserve native catalog display paths and ordering. Keep host filesystem probes and configuration writes restricted to host-compatible paths.
## Testing
Add regression tests for Windows path identity and disabled-skill matching, literal filename characters, catalog resource reads, native path ordering, and foreign-path analytics normalization.
GitOrigin-RevId: c212916dc9fef0e7616542726e107f0d6a313267
## Why
Changing the incremental tools setting on resume should not move tool declarations and base instructions within an existing context window.
## What changed
- Determine the current window's tool declaration mode from its history, and apply the incremental tools setting when building a new window after compaction or a context reset.
- Use that mode consistently for sampling, remote compaction, startup prewarm, prompt debugging, and Guardian input budgeting.
- Preserve tool declarations alongside their comparison baseline in full-history agent forks, including compacted histories; remove them together when rebuilding context.
## Testing
Add request snapshots covering legacy resume, migration after remote compaction or a context reset, disabling incremental tools, and tool updates after restart. Extend agent fork and WebSocket startup prewarm tests to cover retained declarations and incremental tools.
GitOrigin-RevId: 7c46849a1788027a8435082e4d1f7e7f48d2700c
## What changed
Carry `multi_agent_version` from the turn context through resolved configuration facts into turn analytics events.
## Testing
Extend the turn event serialization test to assert that `MultiAgentVersion::V2` is emitted as `"v2"` in `multi_agent_version`.
GitOrigin-RevId: f3e8c98588926d41a18c3db9aa229677e68cf5c1
## Why
Partial answers contain answer fragments while a turn is still in progress. Treating them as final answers defers mailbox delivery and can count empty fragments as empty final responses; filtering them from forked histories and guardian transcripts loses answer content.
## What changed
- Allow mailbox preemption after `partial_answer` messages and keep mailbox delivery open.
- Preserve partial answers in forked agent histories and include them as protected assistant entries in guardian transcripts.
- Exclude empty partial answers from empty-final-response accounting for goals.
## Testing
Add regression coverage for partial-answer phases in live events, sampling continuations, follow-up requests, and forked child requests, plus mailbox preemption, goal accounting, and guardian transcript inclusion.
GitOrigin-RevId: 208c8933e211b42b6c2d0bec6952527b8c9c120e
## Why
Assistant answer text may be followed by more output or tools. A distinct `partial_answer` phase lets callers represent this separately from commentary and a terminal `final_answer`.
## What changed
- Add `MessagePhase::PartialAnswer` and expose `partial_answer` in the app-server schemas and generated TypeScript and Python types.
- Keep the TUI working indicator visible after partial answers and exclude them from the thread history's final-answer summary field.
- Route partial answers to the speakable channel in realtime V3 BEM-tag mode.
## Testing
Add coverage for phase preservation through serialization, input-message conversion, and SSE item start/completion events. Extend TUI snapshot coverage to verify that partial answers keep the working indicator visible while final answers hide it.
GitOrigin-RevId: 1ae3ad9264b6e36ae1cbe6b93df3d345ab4460e6
## What changed
- Record nonempty base instructions as a developer message after tool declarations at the start of each incremental tool context window, and omit the request-level `instructions` field.
- Preserve this prefix when rebuilding history after local or remote compaction, while keeping the compaction summary last.
- Count recorded base instructions once when estimating context usage and trimming tool outputs.
## Testing
Add regression coverage for instruction placement, omission of request-level instructions, compaction ordering, and token accounting. Exercise remote compaction trimming with both request-level and recorded base instructions.
GitOrigin-RevId: 6a6fda92c5c6eb33852a77b38d34c028cf78b164
## What changed
- Accept per-environment `skills.required` lists through `environment/add` and `environments.toml`.
- Before inference, require each named skill to be enabled and supplied by the selected environment. Missing skills fail the turn before a model request, including when environment setup fails.
- Stop applying an environment's requirements after it is deselected. Exempt isolated Guardian reviewers, which have no skill catalog.
## Testing
Add tests for available, missing, disabled, and wrong-environment skills, failed environment setup, recovery after deselection, and propagation of TOML requirements to the correct environment.
GitOrigin-RevId: 302ced4774697d2f414f745fe0535e71f747eda2
## Why
Reviewer compaction can invalidate previously admitted transcript evidence. When a parent checkpoint is available, restarting from that checkpoint preserves the captured parent context instead of relying on a lossy reviewer summary.
## What changed
- Restart an invalidated reviewer once from the action-time parent checkpoint, within the original review deadline.
- Preserve fresh checkpoints during startup and fit review input using the existing budget selection.
- Select transcript evidence from the checkpoint actually loaded by the reviewer. Track the transcript source so cursors remain valid across legacy-history eviction and reset when the source changes.
- Resume transcript deltas after successful recovery, including authorization added after the parent checkpoint.
## Testing
Add regression coverage for checkpoint recovery, post-checkpoint authorization, resumed deltas, and stopping after one restart. Extend history tests to cover both legacy retention and parent-checkpoint reuse, and update the checkpoint-migration snapshot.
GitOrigin-RevId: bc20fb9489afaad963303beff1c79f62fa5023f7
## Why
Manual, pre-turn, and post-turn compaction previously deferred rebuilding initial context until the next turn. Their saved replacement history could omit refreshed instructions and skill catalogs.
## What changed
- Install initial context during local and remote compaction, before the last real user message or summary, keeping the compaction summary last.
- Persist matching full world-state and turn-context baselines with every replacement instead of retaining only partial extension metadata.
- Build replacement context and recompute token usage for the current model even when the previous model performs compaction. Reuse that captured context for eligible current-model retries.
- Capture manual compaction context centrally and propagate preparation errors to the task runner.
## Testing
Update assertions and snapshots for context ordering, refreshed instructions, resume and fork history, current-model checkpoint baselines, and cloud skill catalog retention across compaction and resume.
GitOrigin-RevId: 8572a76c31c6884a3c580b3bbb350d49af876bd1
## What changed
Add `tools_change_count` to turn profiles and turn analytics events. Compare the full model-visible tool list before each sampling request, retaining the previous list across turns and connection resets. The first list establishes a baseline; subsequent changes increment the current turn's count.
## Testing
Add coverage for schema changes, tool additions, ordering changes, empty lists, and tracking across sessions and transport fallback. Extend the turn event serialization test to include `tools_change_count`.
GitOrigin-RevId: 44a15c42dc7d0b35c166666faeb9627b718b33dd
## What changed
When `IncrementalTools` is enabled for Responses Lite, track tool definitions in world state and conversation history. Send the initial catalog once, then append only added or changed definitions while preserving earlier request history. Emit developer notices for removed tools and namespace instruction changes.
Use history-supplied definitions for normal requests and remote compaction, and omit empty `additional_tools` prefixes. Keep the existing tool delivery behavior for the Responses API and when the feature is disabled.
## Testing
Add unit and snapshot coverage for catalog and namespace transitions, empty catalogs, and stable hashing across JSON object key order. Add a request-history scenario covering unchanged tools, a model switch that changes one definition, persisted catalog hashes, and Responses API behavior.
GitOrigin-RevId: 76bde175973b81f8ad23873919da3c8ddcdf6065
yitaoh-oai