Commit Graph
78 Commits
Author SHA1 Message Date
ioanbudea-OAI 0a0b1021f0 Apply elastic Guardian budgets to retained user and sender evidence (#62829)
## Why

Fixed retention limits could discard long user instructions or leave
older sender messages out of Guardian review even when the configured
elastic budget had room for them.

## What changed

- Apply the configured retained-instruction entry limit when capturing
  and restoring user messages. Fixed-budget reviewers use the fixed
  retention window for both user and assistant messages.
- Build elastic sender snapshots at delivery using the receiver's sender
  evidence budget. Give recent user messages first claim on space, then
  include preceding assistant context where it fits. Mark omitted
  evidence and save the snapshot for review and rollback.

## Testing

Add coverage for cross-host delivery after compaction, large sender
snapshots through checkpoint restore and rollback, and how synchronous
and asynchronous reviewers read saved or empty sender snapshots.

GitOrigin-RevId: f0b01a2d558e3286335efac85a4f8645dd9b954b
2026-11-20 20:11:57 +00:00
ioanbudea-OAI ab7a93771e Honor Guardian retention settings for assistant messages (#62828)
## Why

The fixed retention window can discard assistant messages before elastic
Guardian reviews can use them as context for a user's reply.

## What changed

Apply the thread's Guardian retention policy when recording assistant messages,
including confirmed messages sent through the user messaging tool. Keep full
text within the configured record limit and the overall assistant history byte
budget. Preserve an incomplete marker when a confirmed delivery exceeds the
record limit.

Recover the thread's retention settings from saved history and compaction
checkpoints. Fixed-budget reviews continue to use the fixed retention window.

## Testing

Add coverage for record limits during capture and restore, thread-owned
settings recovery, and replay of oversized confirmed deliveries.

GitOrigin-RevId: e828f06824ce6712910a9cc5ef42a438d1ddf27e
2026-11-20 20:10:51 +00:00
yitaoh-oai 4360c5f813 Remove model-selected Search MCP routing (#62808)
## What changed

Use the turn's effective `web_search` mode for standalone and hosted
web search. Remove model-selected Search MCP discovery, fallback, and
provider tracking across thread resumes.

Remove `web_search_provider` from model metadata, the
`tools.client_search_provider` hint from configuration, and the
`hosted_search_policy` feature flag. Update the app-server schemas and
generated SDK types to match.

GitOrigin-RevId: 59e3371f4a43d532a484db52d5998f6770292028
2026-11-20 17:46:32 +00:00
Vivian Fang ea8ccc1d20 Keep explicitly pinned history outside the carryover token budget (#62677)
## Why

An exhausted carryover budget could drop explicitly pinned items, and an
oversized pin could crowd out ordinary user messages during local compaction.

## What changed

Preserve eligible pins in full and in their original order for the current
thread, child threads, or both, according to their retention policy. Apply
the carryover budget to the remaining history, including items retained by
features without an explicit pin.

## Testing

Add regression tests for destination-specific pins at zero budget and for
keeping an ordinary user message alongside an oversized pin during local
compaction.

GitOrigin-RevId: fc8529bcc254d8e7a4a010337ce760e94daa4bef
2026-11-20 05:28:11 +00:00
Vivian Fang a0a30c40e8 Add thread and child retention policies for injected items (#62654)
## Why

The `always` policy ties retention across context windows to inheritance by
spawned agents. Callers need to control those behaviors independently.

## What changed

- `alwaysInThread` carries items across context windows in the current thread
  and excludes them from spawned agents, including full-history forks.
- `alwaysInChildren` passes items still in the parent's context to spawned
  agents and leaves each thread's context-window retention to the usual rules.
- `always` continues to apply in both scopes, including for items saved by
  older versions. Retained tool calls and outputs must be adjacent, with the
  call first and the same retention policy on both items.

## Testing

Extend integration coverage for retention across remote compaction, inheritance
by spawned agents, and injection through the app server.

GitOrigin-RevId: aafede69cfbe149aa52292e6029d248d4bdf6401
2026-11-20 02:59:58 +00:00
Ruslan Nigmatullin 583ec348cf Centralize thread context ID allocation (#62633)
## Why

Response item IDs and context window IDs use separate generator callbacks.
A shared allocator gives history and inference a single API for both.

## What changed

Use `ThreadContextIds` to allocate response item IDs, retained-context
revisions, and compaction window IDs. Move initial window selection into
`InitialHistory` so resumed threads reuse a valid window ID from their
rollout. Give rollout reconstruction its own allocator.

Assign hook prompt IDs through the session's normal response item
recording path.

## Testing

Move the prefixed UUIDv7 item ID test to the allocator crate and add
coverage for JSON escaping in Guardian's item ID token budget.

GitOrigin-RevId: 31232f50648382033506d1f86815ce7c1de78378
2026-11-20 00:44:31 +00:00
Owen Lin 30c69e60a3 Remove cumulative MCP attribution from requests and rollouts (#62589)
## What changed

Stop collecting cumulative MCP tool attribution, sending it in Responses
API request metadata, and writing it to rollout checkpoints.

Use `model_search_state` for search provider history in new rollouts. When
reading older rollouts, decode `mcp_attribution` only to determine whether
Search MCP may have been used. Treat malformed or error attribution as
possible Search MCP use so native search fallback stays disabled.

## Testing

Add a history test that distinguishes legacy Search MCP use, an explicit
record of no use, and missing attribution.

GitOrigin-RevId: a145b5b6b7d2f423d989af10623ef21120f49d4d
2026-11-19 21:06:08 +00:00
Zahan Malkani bd00a8a42e Preserve pending realtime speech when a call ends (#62299)
## Why

A realtime call can end with completed speech that has yet to reach a
Codex turn. The next turn needs that context even if the call's transport
has already disconnected.

## What changed

- For V3 calls with realtime frame history enabled, save pending speech
  as context for the next turn. Restore it on thread reload and exclude
  it from the turn count used by rollback.
- Drain already received transcript events and settle pending handoffs
  before saving the call's closing marker. Report recovery required if
  shutdown cannot confirm that history was saved.
- Keep transport detach separate from ending a call. Reattaching restores
  saved pending speech, and an open call can be ended from persisted
  history after a reload. Use the call ID for TUI attachments so successive
  calls on one thread have distinct history.

## Testing

Add coverage for queued speech at hangup, reconnect and stop after
reload, retries after an uncertain write, and rollback across saved
session-end context.

GitOrigin-RevId: 657491f1d7e446a1a48e22189758810c8bdd82cb
2026-11-18 15:24:29 +00:00
Zahan Malkani ad5dd4a110 Persist realtime handoffs before accepting turn input (#62288)
## Why

Realtime history needs to record which transcript frames reached the
coding agent and which turn received them. If a handoff fails or the
conversation reconnects, Codex needs to distinguish speech that was
already used from speech that is still available for the next input.

## What changed

- Route handoffs through normal turn admission to start or steer a turn.
  When realtime frame history is enabled, persist the transcript frames
  and the receiving turn's input before publishing acceptance or exposing
  the input to the model. Attribute only transcript frames represented
  in the rendered input.
- Reserve transcript frames across overlapping handoffs and release them
  on definite rejection. Retry an uncertain write with the same input
  and require history recovery if the write still cannot be confirmed.
- On V3 reconnect, restore persisted speech that has not been consumed
  and put it before newly captured speech.

## Testing

Add regression tests for attribution after transcript truncation,
overlapping handoffs, V3 reconnect ordering, and preservation of typed
input when a handoff in the same turn fails before model sampling.

GitOrigin-RevId: 432dd51971f7ced20344ebc7f0c7b007bb849f54
2026-11-18 14:26:53 +00:00
Zahan Malkani 4389a526d5 Derive realtime input receipts from thread history (#62251)
## Why

Accepted handoffs and saved transcript chunks need to appear in realtime
history. Their text can come from the same user message that model history
uses, avoiding a second persisted copy for the receipt.

## What changed

- Reconstruct `turnInputAccepted` and transcript chunk receipts from user
  messages annotated with realtime provenance. Use the stored observation
  time when restoring the realtime clock.
- Replace `sessionEndInputSaved` in the app-server protocol with
  `sessionInputSaved` and a `reason` of `chunk` or `sessionEnd`. Update the
  generated schemas and bindings.

## Testing

Add coverage for reconstructing both receipt types, preserving rollout
order in paginated realtime history, and restoring observation time
across a restart after wall-clock rollback.

GitOrigin-RevId: fcb4e0e539f4e4066726e83fa1e945fb659a2ca8
2026-11-18 11:01:02 +00:00
Zahan Malkani 6d9f0199ed Refactor realtime handoff and history access (#62245)
## What changed

Move realtime text routing into `realtime_handoff` and keep it on the shared start-or-steer path. Extract UTF-8-safe range selection for realtime delegation truncation.

Use `RolloutItem` accessors for realtime timestamps and items in recording, clock restoration, and thread history reads. Pass annotated response items to turn-boundary checks in core and migration rollback so callers can work directly with history envelopes.

GitOrigin-RevId: 3dd37cf91ee5abdca84e87b7d7e304749d9baf04
2026-11-18 09:55:30 +00:00
Owen Lin 06789c7618 Preserve web search provider ownership across thread resumes (#62144)
## Why

Search MCP and native web search produce provider-specific references.
Switching providers after a thread resumes can mix those references and
break follow-up calls that use them.

## What changed

Persist whether a thread has used Search MCP, native web search, or
neither. Restore that state from response history, including compacted
history, and use legacy attribution and fallback markers for older
threads. When resumed history lacks a search state or legacy attribution
checkpoint, withhold native fallback unless a legacy marker identifies
native search.

Require every Search MCP call to claim the thread's search provider
before execution, including calls retained by Code Mode.

## Testing

Cover resuming a thread with Search MCP references after discovery
becomes unavailable, both with a saved search state and with imported
history that lacks search checkpoints.

GitOrigin-RevId: 100d21c92b0bdbf737ec36c3474f0bc020df8efc
2026-11-18 00:45:07 +00:00
Zahan Malkani 5943d7acdf Persist timestamped realtime frames before acknowledging completion (#61473)
## What changed

- Record voice transcripts as `frame` items with a `voice_transcript` channel for eligible threads, while retaining legacy transcript support and the existing app-server live and mixed timeline representation.
- Assign first-observed timestamps that never move backward, retain them through completion, and restore the clock floor from history and compaction checkpoints.
- Flush realtime records before completion notifications, including file and initial directory durability barriers for local rollouts.
- Generate a distinct default session ID for each new call, use the call ID when attaching to an existing call, and derive opening record IDs from session IDs.
- Add protocol types for turn-input acceptance and session-end input receipts, excluding those receipts from live notifications and mixed timelines. Update generated schemas and SDK types.

## Testing

Add coverage for frame serialization and required voice payloads, clock rollback and checkpoint restoration, durability barrier failures and retries, per-call session IDs, and persisted transcript availability immediately after completion.

GitOrigin-RevId: 3f6632b517d563861aee085b723475b8945c391c
2026-11-14 06:14:36 +00:00
Hung 5158e9780e Add opt-in file-backed tool audio (#61394)
## What changed

- Add the disabled-by-default `file_backed_tool_audio` feature to prepare fresh function and custom tool audio through the attachment store before persistence. Store returned file IDs in history and preserve them on replay without uploading again.
- Validate and budget mixed text and audio output before uploading retained clips, preserving audio token charges and producer limits. Resolve tool audio durations before continuation compaction checks.
- Keep inline audio on backend upload failures and use text placeholders for invalid or rejected clips. Replace unavailable tool audio files with placeholders in prompts while retaining the original history, so restored files can be used again.
- Treat file-backed tool audio as a competing source when attributing audio or unknown-file errors to user uploads.

## Testing

Add coverage for direct tool and Code Mode audio across follow-up turns, file expiration, restoration, and resume in both legacy and paginated history. Extend tests for truncation before upload, upload failure handling, bounded error placeholders, and error attribution.

GitOrigin-RevId: 66c1e7699c7d92f98c3a2d0f5cecbbd63147eb6c
2026-11-13 23:02:16 +00:00
Vivian Fang c33002259e Add per-item retention policies for injected history (#61344)
## What changed

Allow `thread/inject_items` to accept wrapped items with `retention: "default"` or `retention: "always"`, while preserving support for raw response items. Persist explicit policies separately from legacy client-authored metadata.

Carry explicitly retained items through local and remote compaction, context-window changes, and child-agent creation, subject to retention budgets. Preserve history order and keep retained tool exchanges together; require matching calls and outputs to be adjacent in the injection request and both marked `always`.

Update Quinn dependencies and record exceptions for additional Hickory DNSSEC advisories where DNSSEC is disabled.

## Testing

Update injection tests for explicit retention metadata and extend the compaction scenario to cover a retained tool exchange across repeated compaction and restart.

GitOrigin-RevId: 1195a4ed62dc162a62e58bc2ad733262fbabea27
2026-11-13 18:46:13 +00:00
Vivian Fang 7844815bd7 Add opt-in prompt image sanitization (#61341)
## What changed

- Add the disabled-by-default `safe_image_decoder` feature to decode and re-encode prompt images without source metadata, with byte and decoded-pixel limits.
- Apply JPEG orientation and supported ICC color conversion to sRGB, emitting PNG for PNG/GIF inputs and JPEG for JPEG/WebP inputs.
- Use the selected decoder for image preparation, `view_image` validation, and MCP image validation. Include the decoder in image cache keys.
- Preserve original inline images in history metadata so resume and fork can regenerate prepared images without repeated JPEG encoding changing their bytes.

## Testing

Add regression tests for identical JPEG replay after history serialization and for preventing legacy cache entries from bypassing sanitization.

GitOrigin-RevId: 718cd83d0a233e2eb8efd1a64c14d8944e44c350
2026-11-13 18:44:31 +00:00
Ruslan Nigmatullin 77cbcdb4a6 Preserve thread-local model ID state across replay and compaction (#60942)
## What changed

- Initialize item and context-window ID generators per thread, reuse the stored window on resume, and give history reconstruction a separate item generator.
- Recover item ordinals from checkpoint replacement history, guardian history, and retained context without lowering the cursor established by earlier records.
- Stamp compaction resume metadata with the next item ordinal in append order, tracking the cursor in local and in-memory thread stores.

## Testing

Extend the history ordinal test to verify that a checkpoint records the cursor at its position in a batch, excluding later allocations.

GitOrigin-RevId: 7121b747a039dc5477235078702137d1ff74e789
2026-11-12 00:27:01 +00:00
Ruslan Nigmatullin 9e6d1c5278 Preserve retained assistant revisions across rollout replay (#60839)
## Why

Replaying delivered assistant messages allocates new retained-source revisions instead of preserving those captured in live history.

## What changed

- Capture retained sources when recording assistant deliveries, persist them in rollout metadata, and reuse their revisions during replay.
- Add optional `next_item_ordinal` compaction metadata and a `next_model_id_ordinals` helper to recover item and window allocation positions from checkpoints and subsequent records. Ignore allocation metadata that refers to other threads.

## Testing

Extend tests to compare live and replayed delivery sources, preserve revisions through an older reader's rewrite, and verify ordinal recovery from checkpoints and subsequent local items.

GitOrigin-RevId: cfdeebe8e13b1700b34b9c98696880a90a99ace5
2026-11-11 18:56:47 +00:00
Ruslan Nigmatullin 340a18ed7c Preserve model-facing ID allocations in saved history (#60771)
## Why

Restoring model-facing ID sequences requires allocation ordinals alongside the rendered IDs. Responses API payloads still require string IDs.

## What changed

- Add allocation metadata to `ResponseItemId` while keeping equality, hashing, and ordering based on the rendered ID.
- Preserve optional `id_allocation` metadata in rollout items, inter-agent messages, compaction replacement history, and Guardian checkpoints. Continue accepting legacy string IDs.
- Support structured context-window allocations alongside UUIDv7 strings, validate the `window` salt on restoration, and keep request metadata IDs as strings.

## Testing

Add tests for allocation round trips, legacy ID compatibility, string-only raw response item IDs, and context-window allocation validation.

GitOrigin-RevId: ce1fd964bde9a4e7c527432be99aea87ada30169
2026-11-11 09:58:13 +00:00
Ruslan Nigmatullin 4f8205c808 Use typed context window IDs throughout thread persistence (#60757)
## What changed

Carry `ContextWindowId` through session metadata, compaction checkpoints, rollout recording, and thread creation instead of converting IDs to strings. Add deserialization and JSON schema support while preserving the serialized string format and legacy numeric compaction IDs.

Validate deserialized IDs as UUIDv7 values before restoring context windows, preserving the existing handling of unsupported IDs.

GitOrigin-RevId: fda329d0654f438171a955c9f75550817c32c047
2026-11-11 08:16:52 +00:00
ioanbudea-OAI 2889d8e1bf Normalize turn context permissions to a required profile (#59963)
## What changed

Make `TurnContextItem.permission_profile` required and remove its legacy `sandbox_policy` and `file_system_sandbox_policy` fields. Normalize older histories into a permission profile during deserialization, preserving the restrictive fallback for malformed filesystem rules.

Keep legacy compatibility fields in rollout serialization so older readers remain usable. Update turn recording, metadata extraction, and permission checks to consume the profile directly.

## Testing

Add coverage for legacy normalization, profile-only serialization, missing permissions, and malformed filesystem rules. Extend rollout round-trip fixtures to cover deny rules and filesystem roots from another platform.

GitOrigin-RevId: 27515fec60cccbac5806b98fb71f54e311a09701
2026-11-06 19:29:06 +00:00
jif 48524f337a Preserve pending agent messages across compaction (#59404)
## Why

Agent messages that arrive after a model request starts must survive compaction until a normal inference request includes them. Compactor output alone does not establish that those messages were delivered.

## What changed

- Persist history sequences and sampling checkpoints in harness metadata, capturing the input prefix for each normal inference request.
- Retain pending agent messages within existing size budgets during local and remote compaction, placing the summary before pending messages and subsequent caller input.
- Preserve sampling progress across history replacement and resume, and drain queued mail after interrupted sampling even when compaction leaves context above the token limit.

## Testing

Add regression coverage for request-boundary sampling checkpoints, pending messages across local compaction and repeated remote compaction with resume, and injected messages that interrupt sampling before mid-turn compaction.

GitOrigin-RevId: 26c04ab96b5a1c3cc1d6625ccc74a0f00201d14c
2026-11-04 13:03:01 +00:00
Ruslan Nigmatullin d84c872464 Route response item IDs through a shared generator (#59312)
## What changed

Introduce `ItemIdGenerator` and pass the runtime-owned generator through sessions, model clients, history, hooks, tool outputs, and memory requests. Keep UUIDv7 generation as the default and replace direct calls to `ResponseItemId::new` with the shared callable.

Preserve existing item IDs and backfill missing or empty IDs through the same generator, including forked history before reconstruction and persistence.

## Testing

Extend coverage for UUIDv7 IDs on user messages, agent messages, and additional tools, and assert that forked model context matches the backfilled items persisted to history.

GitOrigin-RevId: 0dd9ccf0f46f0f1005a6acfc740133b03ce82259
2026-11-04 04:02:00 +00:00
Alex Kira a47a518928 Honor Guardian's additional-context setting for tool outputs (#59239)
## What changed

Tag untrusted additional context rendered as tool outputs with history metadata and apply `guardian_include_additional_context` when building Guardian review history. Preserve this metadata across session resume and apply the current session's setting to restored review checkpoints.

Keep ordinary tool outputs in review history and preserve the acting model's additional-context payload.

## Testing

Add coverage for inclusion and exclusion after restart, restored checkpoints after compaction, and asynchronous Guardian review. Verify both line and JSON transcripts and check that metadata does not alter the acting model's request.

GitOrigin-RevId: 8ba270a6aac16398c88db3282e9d666702f462fe
2026-11-03 21:54:52 +00:00
Hung f1b0721c88 Enable uploaded audio with durable rejection recovery (#59234)
## Why

Uploaded audio was rejected even for audio-capable models. Invalid or expired recordings also need to be excluded from later requests so users can continue the conversation.

## What changed

- Accept uploaded audio when the selected model supports audio; otherwise, preserve valid sibling input and report the unsupported attachment before inference.
- Track successful audio decoding by model and provider, and attribute provider rejections against the exact submitted request. Preserve attachments when attribution is ambiguous.
- Persist confirmed rejected file IDs across compaction and resume, omitting them from future requests without rewriting retained user history.
- Check audio availability before remote compaction and retry once without an unambiguously rejected recording. Include guidance to remove or replace affected attachments in audio errors.

## Testing

Add unit and scenario coverage for uploaded audio across follow-up turns, compaction, cold and paginated resume, model switching, ambiguous rejections, expired files, and queued or interrupted input.

GitOrigin-RevId: e701901e7b1c0865c9f2a5918197215f549b06c0
2026-11-03 21:24:16 +00:00
yitaoh-oai e839b828fb Resolve model-selected search providers with policy-aware fallback (#59036)
## 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
2026-11-03 02:08:56 +00:00
Hung 2546265e3e Add audio file references with explicit runtime rejection (#58430)
## 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
2026-10-30 20:53:25 +00:00
Steve Coffey f088e0c4ff Allow replacing dynamic tools between model steps (#57553)
## 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
2026-10-27 19:50:41 +00:00
Charlie Marsh c7234c1174 Box session metadata and compaction payloads in RolloutItem (#55733)
## What changed

Store `RolloutItem::SessionMeta` and `RolloutItem::Compacted` payloads in `Box` values and update their producers, consumers, and test fixtures. Preserve the existing serialized rollout format.

GitOrigin-RevId: 59fcc4beb47d23160872dcb6e95ace9b142d3d11
2026-10-21 02:41:05 +00:00
pakrym-oai 5ef96ab278 Add opt-in output token replay for OpenAI requests (#52742)
## 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
2026-10-10 01:25:00 +00:00
Won Park 04ad786998 Add retention annotations to submitted tool outputs (#52671)
## What changed

Add `TurnInput::AnnotatedResponseItem` with a caller-supplied `retain` annotation. For function-call outputs, `retain: true` sets `client_authored` history metadata while keeping annotations out of model requests. Accept annotated items through the existing turn submission paths, including continuation.

## Testing

Add an integration test covering retention metadata during turn start, steering, idle submission, and rollout storage. Verify that ordinary outputs remain unmarked and model requests contain neither `retain` nor `client_authored`.

GitOrigin-RevId: 7285dde31149b17b0dcf31d4ea1c7cba66d9efe6
2026-10-09 18:35:53 +00:00
Adam Perry @ OpenAI 780d7ab097 Track history initialization in Responses turn metadata (#52325)
## What changed

Add `history_initialization` to `x-codex-turn-metadata`, distinguishing `new`, `cleared`, `cold_resume`, `warm_fork`, `cold_fork`, and `supplied_history`.

Classify forks by whether the source runtime is loaded and has not shut down. Preserve the classification for the runtime's lifetime, including warm resumes and compaction. This records how history was initialized; it does not establish whether tool metadata is missing. Reserve the field against client-supplied overrides.

## Testing

Extend coverage for warm and cold forks, cold resumes, warm reconnects, supplied history, WebSocket metadata serialization, and reserved-field filtering.

GitOrigin-RevId: 468153512a84d0f4d4ea742198a29e0fa3873a72
2026-10-08 23:34:16 +00:00
Eddie Chen 551bd409eb Preserve turn attribution across recovery and compaction (#51402)
## Why

Recovering an interrupted turn replaced its original trigger with `retry` and restored only its root lineage from model context. Turns suspended before that context was written could lose their attribution entirely.

## What changed

- Persist regular-turn attribution in `TurnStartedEvent` and compaction resume metadata, including the trigger, parent and root turn IDs, and initiating agent path.
- Restore matching attribution during recovery, keep unknown triggers absent, and retain the model-context root fallback for older records.
- Reconstruct attribution from surviving turns across rollback and late completion or abort events, and clear inherited attribution when spawning a forked agent.

## Testing

Add coverage for repeated cold recovery, suspension before context persistence, legacy rollouts, fresh attribution on subsequent turns, compaction checkpoints, rollback, and late terminal events.

GitOrigin-RevId: 85b5763e8fbd48a1ea2c07cfda8196e63c2cd33a
2026-10-06 15:50:27 +00:00
Owen Lin c73775f19e Remove initial messages from session configuration events (#50360)
## What changed

Remove `initial_messages` from `SessionConfiguredEvent` and the corresponding schema. Stop attaching replay history to startup events for resumed or rejoined threads, and remove `InitialHistory::get_event_msgs`.

## Testing

Update the MCP resume test to verify persisted tool events through `load_history`, and retain coverage that resumed requests include prior conversation items.

GitOrigin-RevId: 54b72bc44423c9b1ef2051891ef392c0ede32fe5
2026-10-02 16:55:41 +00:00
felixxia-oai 5ca55db09c Deduplicate Guardian retained-context omission notices (#49796)
## Why

Guardian review continuations can repeat retained-instruction omission notices even when availability has not changed. When retained context becomes complete, earlier notices also need to be superseded explicitly.

## What changed

Track delivered user-instruction and assistant-context omission states in host metadata scoped to the retained-instructions section. Suppress redundant notice-only sections when the state is unchanged, and emit an availability-change notice when an earlier omission clears, including when no retained originals remain.

Treat missing delivery metadata as unknown. Matching notice text in coalesced tool output cannot establish host delivery, and incomplete history entries discard omission delivery proof.

## Testing

Add coverage for repeated notices, independent histories, availability recovery, and matching notice text in multimodal REPL output. Extend app-server review-continuation tests to assert that each effective classifier context contains the retained user-instruction omission notice exactly once.

GitOrigin-RevId: c6d52dd1dafb284af34b46174aecc7c8cc9b8404
2026-10-01 01:13:53 +00:00
felixxia-oai 47a8bd7321 Avoid duplicate sync reviews in Guardian classifier continuations (#49795)
## Why

Retained Guardian classifier conversations already contain previously delivered synchronous reviews. Appending those reviews again repeats evidence and consumes input budget.

## What changed

Track each review completion with a stable identity in host-only history metadata. Filter already delivered reviews before composing and budgeting continuation input, while preserving distinct completions even when their text is identical.

Fresh conversations receive the review evidence again, and shortened messages no longer count as complete delivery. Delivery metadata stays out of model requests.

## Testing

Add coverage for delivering retained reviews once, appending newly completed reviews, restoring evidence after failed streams, and retaining reviews when history lacks complete delivery metadata.

GitOrigin-RevId: b8d58ffdd7d9c484676af707c65fa50020aae73c
2026-10-01 01:13:29 +00:00
felixxia-oai 92bc601ad6 Reuse unchanged history snapshots when resuming threads (#49600)
## Why

Resuming a thread rereads persisted history even when a preloaded snapshot is unchanged. Reusing that snapshot must still account for writes committed before writer ownership is acquired.

## What changed

- Carry optional, non-serialized revisions from stored history and model context through thread and agent resume paths.
- Validate the revision under writer ownership and reuse the shared snapshot when it matches. Reload stored history when the revision is missing or changed, while preserving explicit history overrides.
- Resolve the selected rollout before loading context, validate compressed snapshots before materialization, and enforce archived-thread restrictions on canonical paths.

## Testing

Extend coverage for snapshot reuse and late-write refresh in legacy and paginated histories, compressed rollout reuse, archived rollout symlinks, and cold child restoration that includes edits made after preloading history.

GitOrigin-RevId: e1fd7aee6ef2b30f41951e6b7e6b01c3e1420cab
2026-09-30 11:59:13 +00:00
felixxia-oai 21eb35513d Preserve independent Guardian history across parent compaction (#48779)
## Why

Guardian reviews need to retain original evidence across parent compaction when checkpoint reuse is disabled, including after resume and rollback.

## What changed

- Make disabling `guardian_reuse_parent_compaction` select a bounded, independent review transcript while preserving thread-owned authorization.
- Keep synchronous reviews and asynchronous scoring independent of parent checkpoints, and allow reviewer sessions to continue using transcript deltas across parent compaction.
- Persist transcript entries with rollback provenance while remaining compatible with older checkpoint readers and metadata-free checkpoints.
- Exclude compaction output from the transcript and synthetic summaries from rollback turn boundaries. Use acceptance ordering to remove rolled-back evidence even when persistence order differs.

## Testing

Add regression coverage for reviewer continuity, evidence retention across compaction and resume, rollback after local and remote compaction, and checkpoint serialization compatibility.

GitOrigin-RevId: fea0852cc5537372854f4a34bacb6bfda6d13e88
2026-09-27 20:15:23 +00:00
Ankush Gupta 88235f881d Retain confirmed Code Mode messages for Guardian reviews (#48725)
## Why

Guardian needs the assistant's delivered question to interpret a user's reply. Messages sent through nested Code Mode tools must remain available as review context, and approvals based on earlier context must become stale when a new delivery is confirmed. Assistant messages do not themselves grant authorization.

## What changed

- Capture successful hosted User Messaging sends after input rewriting and before result callbacks or post-tool hooks.
- Retain bounded delivery text in acceptance order, persist it through cancellation and shutdown, and preserve it across compaction and resume with an older-client-compatible rollout encoding.
- Track assistant review context separately from user authorization, invalidating pending and cached approvals for both local and worker reviews when that context changes.
- Associate retained deliveries with the correct instruction or communication boundary during rollback.

## Testing

Add coverage for nested messaging review context, fast replies and communication ordering, shutdown persistence, stale local and worker approvals, rollout compatibility, and rollback retention.

GitOrigin-RevId: 1bc1ec9e8947560b99f4b39a42ddb9f2099f519b
2026-09-27 15:03:38 +00:00
jamy-OAI 5f3180c793 Preserve model and access program pairs during compaction (#48224)
## Why

Compaction using the previous model can inherit the current turn's access program, producing a model/program pair that the server rejects.

## What changed

- Persist `cyber_access_program` in previous-turn settings and restore it during rollout reconstruction.
- Use the previous turn's access program when compacting with its model, preserving an absent program instead of inheriting the current selection.
- Pass the selected access program into local compaction prompts.

## Testing

Add regression coverage for local and remote compaction after model switches, including resume, fork, rollback, and compaction checkpoints. Verify fallback and subsequent sampling use the current model/program pair, and API-key sessions do not inherit access-program authorization.

GitOrigin-RevId: eaa7162e8711446dc9eb1bfff194e711abd54d7e
2026-09-25 22:14:19 +00:00
olliem-oai a0b85c7a66 Fix Guardian retained context spacing and empty assistant handling (#48158)
## Why

Complete assistant messages with empty text can produce misleading omitted-context notices in Guardian reviews.

## What changed

- Skip complete, empty assistant messages when recording and rendering retained context, while preserving omission notices for incomplete messages.
- Separate retained instruction fragments with blank lines and keep blank lines free of role prefixes, while labeling every nonempty line with its original role.
- Update source-order guidance detection and empty-section deduplication for the added spacing.

## Testing

Extend retained-context tests to cover empty completed messages across resume, restored empty messages, incomplete-message notices, blank-line spacing, and role labeling. Update Guardian prompt snapshots.

GitOrigin-RevId: e945717b18689fa4538d9b617b89a644858ab1c8
2026-09-25 18:28:46 +00:00
felixxia-oai d5cbfe1455 Deduplicate retained instructions against Guardian transcripts (#48109)
## Why

Synchronous Guardian reviews can include the same user instruction in both the retained instructions section and the conversation transcript. Avoid this duplication while preserving complete authorization evidence and its original acceptance order.

## What changed

- Carry retained source metadata into transcript rendering. Omit a retained instruction when the transcript delivers its complete matching source revision.
- Label qualifying transcript entries with retained source order and protect them from budget truncation. Keep separate retained instructions when source metadata is missing or incomplete.
- Track delivery of source-order guidance so it remains available even when all retained instructions are delivered through the transcript. Invalidate that delivery proof when messages are shortened.

## Testing

Add coverage for complete-source deduplication, missing or incomplete metadata, budget protection, guidance redelivery, and source metadata preservation across resume. Update Guardian review assertions and snapshots to verify instructions are delivered once.

GitOrigin-RevId: 20ad0304501fa49f1596c4cfaf3ef89d64815753
2026-09-25 14:54:51 +00:00
felixxia-oai b35a7afbe8 Preserve recent authorization context for Guardian reviews (#48098)
## Why

User messages could fill the root context limit before assistant questions were considered, while commentary could displace relevant conversational context. Guardian needs that context to interpret user authorization when reviewing subagent actions.

## What changed

- Select recent user and assistant evidence together under a shared message limit, discarding assistants with unknown ordering first.
- Exclude explicit assistant commentary and retain message phases across compaction. For older retained records, check available source messages for commentary phases.
- Mark missing instructions or assistant context when the shared limit removes corresponding evidence, and use remaining space for legacy instructions.

## Testing

Update Guardian authorization tests to cover shared message limits, commentary filtering across compaction and resume, and preservation of ordered replies when legacy assistant context has unknown ordering.

GitOrigin-RevId: c369a2760917cba9d7dfca37fb5e85a7ec49012f
2026-09-25 14:06:57 +00:00
felixxia-oai 86be5320b0 Deduplicate retained instructions across Guardian reviews (#48060)
## Why

Reused Guardian sessions already contain previously delivered instructions. Follow-up reviews can avoid repeating that evidence, but must deliver it again if it changes or compaction removes it.

## What changed

- Track retained source identity, revision, and completeness in host-owned history metadata, preserving revisions during replay.
- Omit retained instructions from synchronous review inputs only when the same complete revision remains in reviewer history.
- Restore missing evidence after reviewer compaction, enforcing the request budget before persisting additions. Classify restored evidence as context rather than new user authorization.
- Keep source-order labels stable across eviction and retain full delivery for legacy records without reliable provenance.

## Testing

Add coverage for consecutive-review deduplication, changed revisions, missing or incomplete metadata, and restoring compacted evidence exactly once. Extend compaction and resume coverage to verify retained revisions.

GitOrigin-RevId: e90fca16cde0a0970da36dd873af2ca1dbaaf4d3
2026-09-25 10:49:50 +00:00
peilin-openai 62197a65e5 Add diagnostic reasons to MCP attribution errors (#47899)
## Why

The `attribution_error` status does not explain why MCP attribution is incomplete. Preserve a diagnostic reason so callers can distinguish history and recorder errors from request serialization failures.

## What changed

- Add optional `error_reason` diagnostics to `McpAttribution`, retaining the first recorder error through checkpoints, compaction, resume, and fork.
- Identify missing or invalid checkpoints, conflicting sources, invalid sources, and poisoned recorder state. Use `restored_error_unknown` for older error checkpoints without a reason.
- Distinguish `payload_too_large` from `serialization_failed` in request metadata fallbacks.
- Tolerate missing, unknown, or malformed diagnostic reasons when reading history, without letting diagnostics determine attribution status.

## Testing

Add coverage for reason preservation, recorder poisoning, backward-compatible checkpoint decoding, resume and cold-fork restoration, and request metadata serialization, including oversized payloads.

GitOrigin-RevId: 389b9446fae119f793978f3e7e4a926677e84d8c
2026-09-24 18:06:07 +00:00
felixxia-oai e7f119819a Preserve human overrides across repeated heartbeat instructions (#47851)
## Why

Unchanged scheduled heartbeat instructions should not become fresh authorization evidence that supersedes a later human override or floods Guardian review context.

## What changed

- Capture heartbeat provenance per input, including when steering an active turn, and persist it in history and retained context.
- Coalesce unchanged instructions for each automation at their original acceptance order. Preserve changed instruction versions and ordinary human messages.
- Represent repeated heartbeats as references to earlier instructions in Guardian transcripts, while tracking the latest input turn separately for trusted skill selection.
- Limit coalescing to recognized, bounded heartbeat envelopes with host-provided provenance.

## Testing

Add regression coverage for human overrides across repeated and steered heartbeats, instruction changes, checkpoint restoration, transcript references, envelope parsing, and latest-turn trusted skills in delegated reviews.

GitOrigin-RevId: 31e444ddb685a1a618be26d3cc32fb562296946b
2026-09-24 14:48:41 +00:00
felixxia-oai a6df7385e6 Make thread-owned Guardian context always enabled (#47686)
## What changed

- Deprecate and ignore `features.guardianv2.thread_context`, including profile overrides, with a migration notice. Ignore managed requirements for the removed flag with a warning.
- Preserve surviving legacy root instructions across compaction and resume when retained instructions are incomplete. Include unmatched instructions in worker authorization context with an explicit unknown-ordering notice, without replacing newer retained facts or expanding the reviewer's model window.
- Use surviving source calls to roll back verified answers from older checkpoints that lack acceptance-order metadata.

## Testing

Add regression coverage for legacy instruction recovery, compaction and resume, source matching, and answer rollback around same-turn user input. Update configuration and Guardian review tests to verify that the deprecated opt-out still uses thread-owned context and rejects incompatible compaction evidence.

GitOrigin-RevId: a65936c9e0b912e7667d1385a2ec53df726565de
2026-09-23 22:44:59 +00:00
felixxia-oai 29aa1df62e Retain assistant context for Guardian authorization reviews (#47582)
## Why

Ordinary user replies can depend on preceding assistant messages. Losing those messages during compaction can leave Guardian without the context needed to interpret a reply's scope.

## What changed

- Retain original assistant messages, including commentary, with their recorded order across checkpoints and resume. Exclude compaction output from this evidence.
- Include assistant context in local and delegated reviews as untrusted context, with separate storage limits and lower budget priority than user instructions.
- Prefer assistant context preceding ordinary replies when selecting bounded root evidence, and report unavailable context without inferring what a reply refers to.
- Preserve inherited context scope and remove assistant evidence beyond rollback boundaries, including during rollout migration.

## Testing

Extend coverage for compaction and resume, checkpoint migration, queued-input ordering and rollback, role labeling, and evidence budgets that preserve user restrictions while dropping assistant context.

GitOrigin-RevId: 886189b0be29ffc3d324b56f5d0b673d6b8d84da
2026-09-23 16:15:55 +00:00
Owen Lin e7bbc79f48 Resume model context from the latest compaction boundary (#47365)
## Why

Resuming a compacted thread should use the compaction and its newer records without depending on older turn contexts or restoring stale state from before the compaction.

## What changed

- Stop paginated model-context scans at the newest compaction with `replacement_history` and `window_number`. Fall back to full replay if the newest compaction lacks either field.
- Align reconstruction with that boundary and restore `last_started_turn_id`, previous turn settings, and the multi-agent version from compaction resume metadata. Allow newer completed turns to override the saved settings.
- Keep canonical session metadata alongside the bounded replay and use the same scan boundary for subagent rollout migration.

## Testing

Add coverage for cold resume after mid-turn compaction, bounded/full replay equivalence, missing companion state, settings precedence, and incomplete-compaction fallback.

GitOrigin-RevId: f836464bb5cadd9be4a4e05b5d1679223713fc87
2026-09-22 21:17:08 +00:00
Owen Lin 286d4ecf44 Persist resume metadata in compaction checkpoints (#47323)
## Why

Compaction companion records do not represent all resume state, including the last started turn and previous turn settings. The selected multi-agent runtime also needs a persisted value when no companion turn context is present.

## What changed

- Add optional `resume_metadata` to `CompactedItem`, recording the multi-agent version, last started turn ID, and previous turn settings. Preserve the distinction between absent metadata and explicitly empty values while retaining compatibility with older records.
- Capture this metadata under the settings persistence lock alongside the compaction checkpoint.
- Rewrite inherited metadata to use the child agent's runtime, and clear previous turn settings when the child must rebuild context.
- Clear pending continuation state before delivering acceptance of standalone settings updates, while holding the persistence lock.

## Testing

Add coverage for metadata round trips with empty values, legacy reader compatibility, checkpoint companion records, inherited fork metadata, and continuation invalidation while settings event delivery is blocked.

GitOrigin-RevId: f30b7feaa7e2fdb5b65fc3c53b37fd7a7fe27903
2026-09-22 16:35:30 +00:00