## What changed
- Store the shared `GuardianReviewSessionManager` in thread extension data and let the Guardian extension spawn and fork reviewers through `ThreadManager`, independently of async scorer startup.
- Preserve reviewer reuse, concurrent forks, cancellation, inherited environments, and user instructions. Gate managed reviewer startup on parent registration and remove terminated reviewers from the manager.
- Retain the `guardian` subagent identity in saved conversations. Reject resume of live Guardian reviewers and reject client archive or delete of live owner-managed workers with JSON-RPC error `-32600`; allow normal access after the owner releases them.
- Keep standalone reviewer support for inline delegates and remove duplicated extension prompt and configuration code.
## Testing
Add app-server tests covering reviewer reuse and concurrent forks, inherited instructions and prompt cache keys, live resume and removal rejection, saved reviewer discovery and resume after parent shutdown, and strict Guardian assessment inside an inline review.
GitOrigin-RevId: 4f9893c1ade151eacb4c11ffeae3d9e9ec019187
## Why
Device-authenticated verification needs an app response outside automated approval and review. Previously, MCP user-verification requests were always cancelled.
## What changed
- Add typed `openai/userVerification` elicitations carrying a title, description, and challenge through core and app-server.
- Route verification to one connection enabled by trusted host activation. Restrict responses to that owner, exclude requests from replay, and cancel on disconnect or authentication changes, including account switches away and back.
- Return proofs in `content`, discard response metadata, and cancel malformed acceptances. Suppress verification response logging and analytics, and send only a generic verification notice to realtime.
- Keep the mode experimental and exclude it from stable schema exports. Capability advertisement remains disabled, and the TUI cancels verification requests.
## Testing
Add coverage for owner isolation, disconnect and authentication races, proof response handling, automated-review bypass, inactive-host cancellation, stable schema filtering, and realtime payload exclusion. Update the MCP integration test to verify that an activated request holds the tool until the client responds.
GitOrigin-RevId: c8c7b6691a3f7756682bbcca9def1df0d1feb906
## Why
Shell Snapshot V2 previously captured the login-shell environment lazily when a
command started, adding that work to the command path.
## What changed
- Start one asynchronous snapshot capture after turn hooks accept the turn, so
eligible local `exec_command` calls can reuse it without blocking the model.
- Use the turn's resolved tools, permissions, sandbox, environment, and shell
configuration, and skip prewarming for unsupported or network-managed cases.
- Cancel outstanding prewarming during shutdown. Keep failed speculative
captures retryable by real commands without consuming their retry budget.
## Testing
Added coverage for non-blocking capture and reuse, failure recovery, sandbox and
permission isolation, shutdown cancellation, hook-stopped turns, eligibility
gates, and concurrent capture retries.
GitOrigin-RevId: ff24a4f0328c06653e09d03fd0a3db1eccd65030
## Why
Context compaction can replace original conversation items that Guardian still needs to review approval-sensitive actions.
## What changed
- Retain a bounded, chronological review history independently of the model's compacted history.
- Keep separate limits for user messages and other transcript items so tool traffic cannot evict user instructions.
- Use the retained evidence for synchronous and asynchronous Guardian transcripts, while resetting it after rollback or history reconstruction.
## Testing
- Cover retention through compaction and eviction, reset behavior after rollback, independent size limits, and asynchronous transcript construction.
GitOrigin-RevId: c5851e992820504fc18de8cfec7f6f9f1b69f316
## What changed
- Add an optional `timeoutMs` parameter to `thread/shellCommand` and propagate it to user shell execution.
- Preserve the one-hour default when the parameter is omitted or `null`, allow longer deadlines, and treat `0` as an immediate timeout.
- Reject negative or otherwise invalid values before execution. Timing out an auxiliary shell command leaves its active turn running.
## Testing
- Cover default, extended, immediate, invalid, standalone, active-turn, and interruption behavior.
GitOrigin-RevId: d72f67793d9766b7383f28d30140270a4a76a495
## Why
App-server clients need the explanation and steering instruction returned with a
misalignment policy violation to offer a user-confirmed continuation. Missing or
incomplete details must continue to behave as a terminal block.
## What changed
- Parse optional misalignment classification, explanation, and steering details
from streamed and HTTP Responses errors and propagate them through core errors.
- Include the details in live app-server `error` and `turn/completed` payloads and
export the corresponding protocol schemas and TypeScript types.
- Keep explanations and steering messages out of serialized rollout events and
redact them from debug output.
- Document how clients can resume with `turn/start` after user confirmation.
## Testing
- Cover streamed, HTTP, and WebSocket-wrapped errors, including malformed and
classification-only details.
- Verify live app-server notifications expose resumable details without writing
the explanation or steering message to the rollout.
GitOrigin-RevId: 329258a444c2cd91d0c57ab1720830b33ddcfac5
## What changed
- Add the experimental `turn/start.cyberAccessProgram` option with `standard`,
`daybreakBlue`, and `daybreakRed` values.
- Forward the selection as `access_programs.cyber` on Responses, WebSocket,
and remote-compaction requests made with ChatGPT authentication, while
omitting it for API-key and custom-provider requests.
- Preserve the per-turn selection across recovery, compaction, and child-agent
turns without making it a persistent thread setting.
## Testing
- Cover app-server forwarding, authentication boundaries, WebSocket reuse,
turn recovery, compaction, and child-agent inheritance.
GitOrigin-RevId: d2eb468f365b2214c5099bc21741cf30f8bd2eb5
## What changed
- Add a feature-gated `TurnSettings` operation that can update the model,
reasoning effort, reasoning summary, and service tier for subsequent steps of
a named running turn without changing future thread settings.
- Report whether an update was applied, rejected, or lost its live target.
- Revalidate managed constraints and preserve the turn's admitted approval and
Guardian safety properties before publishing an updated settings snapshot.
## Testing
- Cover sparse updates, step capture and ordering, target replacement, managed
policy changes, model metadata safety, and separation from future settings.
GitOrigin-RevId: ecaaaa95b4fd4ec9d62265bef551582dddb36d78
## Why
Thread settings can change while a turn is running, and delegated review or
compaction steps can select a different model. Each request needs a consistent
set of settings resolved against the model that will execute it.
## What changed
- Capture immutable resolved settings for each model step, including model
metadata, reasoning options, service tier, approvals, and personality.
- Keep in-flight steps on their captured settings while applying thread updates
to subsequent work.
- Resolve inherited reasoning summaries and filter service tiers against the
selected model and feature configuration for review and compaction steps.
## Testing
Added coverage for settings updates during paused turns, model changes,
previous-model compaction, review-model defaults, and service-tier filtering.
GitOrigin-RevId: fca4b3b305379f1bf26a5cac75ee0dd1747ac1f6
## What changed
- Represent compaction summaries and Guardian-approved actions as typed contextual fragments.
- Inject subagent notifications through the same fragment path.
- Annotate the resulting request items with `compaction.summary`, `guardian.approved_action`, and `multi_agent.subagent_notification` content kinds while preserving their existing roles and text.
## Testing
- Verify the content-kind metadata on compacted summaries, approved actions, and subagent notifications.
GitOrigin-RevId: 01c93e7f08b2dc1c1e85ba69e82bfb5f56e6cca5
## Why
An active root turn needs to be stopped without marking it complete or aborted
before another runtime can recover the same turn ID.
## What changed
- Add `CodexThread::suspend_turn_and_shutdown` and `SuspendTurnOutcome`.
- Flush history, stop the active regular task, close the history writer, and
shut down the session without recording a terminal turn event.
- Reject suspension when no supported turn is active or the loaded agent
subtree still contains a live descendant.
## Testing
- Verify that suspension preserves unfinished history and allows the turn to be
recovered under its original ID after the descendant guard is cleared.
GitOrigin-RevId: 1f9b019d07c51474ec2d991d263bc15cdd4f89ad
## What changed
- Capture accepted, successful `node_repl` results from Code Mode and include them as bounded, untrusted evidence in Guardian review prompts.
- Enable the enhanced transcript through `guardian_enhanced_node_repl_transcripts` or when the selected model requires automatic Node REPL review.
- Keep this evidence out of the parent model history, exclude encrypted and failed results, and avoid resending admitted evidence when a Guardian session is reused.
- Clear retained evidence and invalidate review sessions when a thread is rolled back.
## Testing
- Add unit coverage for evidence ordering, escaping, truncation, and empty responses.
- Add an integration test covering feature-disabled, feature-enabled, and model-required behavior, including filtering and review-session reuse.
GitOrigin-RevId: edc3dca5d5b88d472a492f35531ec46889a89d72
## What changed
- Add `RecoverTurnRequest` and `CodexThread::recover_turn_if_idle` to resume an interrupted regular turn with its existing turn ID and updated thread settings.
- Treat recovery separately from automatic idle work so it can resume in Plan mode without injecting an empty user message.
- Reject recovery while another turn is active without applying settings or queuing input.
## Testing
- Cover successful Plan-mode recovery, turn ID preservation, and the absence of an empty user message.
- Cover active-turn rejection and verify that it leaves settings and pending input unchanged.
GitOrigin-RevId: 2376fdc78b5d48b571633981960741ba269ab1ef
## What changed
- Add `TurnInputRequest` and typed submission results for atomically starting a turn, steering the active turn, or declining input with a specific reason.
- Expose `start_or_steer_turn`, `start_turn_if_idle`, and `steer_turn` on `CodexThread`, and migrate Core consumers to these APIs.
- Make app-server `turn/start` steer an active regular turn and return that turn's ID. Reject incompatible output schemas and non-steerable turns without applying settings or enqueueing input.
## Testing
- Cover concurrent start-or-steer submissions, accepted and rejected settings updates, output-schema compatibility, idle-start rejection, and app-server steering.
GitOrigin-RevId: dd9b5528d76ec650c019e97af420bc13190ea86a
## What changed
- Mark developer messages received from clients with `client_authored` harness
metadata when `retain_client_developer_messages` is enabled.
- Carry that provenance through pending input and persist it alongside rollout
response items without adding it to provider requests or the serialized
`TurnInput` payload.
- Preserve existing behavior for unannotated response items and when the
feature is disabled.
## Testing
- Cover client item injection before and during turns, additional application
context, rollout replay, image resize notices, and serialization boundaries.
GitOrigin-RevId: f385c177732becde732272e4b8440eb91f5c1039
## Why
`parent_turn_id` identifies the immediate caller of a nested request, but not the
top-level turn that initiated a multi-level chain of work.
## What changed
- Add `root_turn_id` to turn state, submissions, and Responses API metadata.
- Propagate the root through spawned agents, delegated work, review sessions, and
turn-triggering inter-agent messages.
- Omit the root when background work has no user turn or combined input makes the
root ambiguous.
## Testing
Added coverage for nested agent resumes, reviews, mailbox delivery, steered input,
and reserved metadata handling.
GitOrigin-RevId: 60057db2ee81f4efefa946a9f63c74ff98b93929
## What changed
- Add `ResponseItemEnvelope` as the history-owned wrapper for a `ResponseItem`.
- Store envelopes in `ContextManager` while keeping raw-item iteration and consuming APIs for existing callers.
- Preserve envelopes when compaction and history normalization rewrite retained items, and wrap newly synthesized items.
## Testing
- Cover envelope access, mutation, and conversion back to raw response items.
- Cover annotated history replacement and annotated user-message collection for compaction.
GitOrigin-RevId: 3fa2e7bdd981b08cc7aec410c80deccaecdabab4
## What changed
- Add `PersistContext` to the thread-store persistence contract so stores can
identify persistence requested immediately before model sampling. A
`TurnStart` request may be enqueued in the background when later flush and
shutdown operations fence it and report any failure; all other persistence
remains synchronous.
- Use the turn-start context only for the initial input of a regular turn, while
admission acknowledgments, steered input, and other persistence paths keep
the standard durability behavior.
- Flush pending metadata before turn-start persistence and ensure shutdown is
still attempted if that metadata update fails.
## Testing
- Verify that developer instructions and user input are persisted before the
model request completes for a default paginated-history thread.
GitOrigin-RevId: 162f0ec796a61aebe66ca0b909fbcccec3047b85
## What changed
- Remove `Clone` from `Submission` and remove `Clone` and `PartialEq` from `Op`.
- Consume operations directly in the submission loop instead of cloning them.
- Limit thread-manager test capture to the operation variants under test and update assertions to compare their relevant fields.
GitOrigin-RevId: 781cc631391b8583d94dad5ca45f589856c0b21a
## What changed
- Add `codex-history` for model-history and persisted-rollout domain types, including `RolloutItem`, `RolloutLine`, `CompactedItem`, and initial/resumed history state.
- Re-export the persisted types from `codex-rollout` and update consumers to use the new crate boundary instead of `codex-protocol`.
- Preserve existing rollout serialization, including legacy numeric compacted-window IDs.
## Testing
- Add `codex-history` tests for rollout JSON round trips, compacted-history compatibility, persisted history modes, and multi-agent version selection.
GitOrigin-RevId: 944daa9297ddd231d3aebbdcb05fff4adf8b4e1b
## Why
Hook configurations can mark command handlers as asynchronous, but Codex previously skipped those handlers outside `SessionEnd`.
## What changed
- Run asynchronous command hooks in the background with a per-session concurrency limit, while keeping `SessionEnd` hooks synchronous.
- Prevent asynchronous hooks from blocking, stopping, rewriting, or otherwise controlling the operation that launched them.
- Deliver warnings and additional context at safe turn boundaries: inject results into an active turn after sampling, or buffer them ahead of the next user prompt when the session is idle.
- Preserve in-flight hooks across configuration reloads, scope spilled output to the thread, and abort outstanding work during session shutdown.
## Testing
Add unit and integration coverage for background scheduling, concurrency, output parsing, active and idle result delivery, configuration reloads, and shutdown cleanup.
GitOrigin-RevId: 8094552e4afe7b47b09a61bb575bb20f4e491d8d
## What changed
- Track approved command prefixes separately from the stable permissions
instructions in world-state snapshots.
- Emit only newly approved prefixes after an exec-policy amendment instead of
appending the full permissions block again.
- Preserve prefix updates when full permissions instructions are disabled and
remain compatible with legacy world-state snapshots.
## Testing
- Cover incremental prefix additions, removals, legacy snapshots, history
rollback, and approval flows with permissions instructions enabled or disabled.
GitOrigin-RevId: e6f68c6a91be82750e70e28456f2b7c58607fbd8
## What changed
- Add `CodexThread::submit_user_input_and_wait_for_admission` and export the
`UserMessageAdmission` result through `codex-core-api`.
- Resolve submissions only after they start a new turn or steer the active
turn, returning the accepting turn ID in either case.
- Return errors for invalid operations, rejected thread settings, and session
termination instead of leaving admission waiters unresolved.
## Testing
Add integration coverage for concurrent start-and-steer submissions, settings
rejection and recovery, non-user operations, and submission after shutdown.
GitOrigin-RevId: 21f1ede38d399baa9836d61a59b9f9af2dbf83b8
## What changed
- Propagate the initiating turn ID through agent spawns, follow-up tasks, reviews, and delegated Codex sessions.
- Add `parent_turn_id` to Responses client and turn metadata while keeping it out of external MCP metadata.
- Preserve parent-turn provenance across queued agent messages when their triggering parent is unambiguous.
## Testing
- Cover spawned, resumed, nested, reviewed, delegated, and WebSocket request metadata.
- Verify queued messages do not claim ambiguous or queue-only parent turns.
GitOrigin-RevId: 481fdebbe7df2031880fe259509273cce50b20a8
## What changed
- Update the `rmcp` model and transport integrations for `3.0.0-beta.3`.
- Preserve legacy `elicitation/create` form requests, schema defaults, metadata, and wire-format compatibility while adopting the new elicitation types.
- Harden OAuth discovery by using GET-first discovery without starting an MCP session, preventing cross-origin header redirects, validating authorization-server and callback issuers, and retaining compatibility with metadata that omits an issuer.
- Keep stored OAuth credentials readable and avoid requiring reauthorization after transient refresh failures.
## Testing
- Add coverage for legacy elicitation round trips, OAuth discovery variants, callback issuer validation, credential migration, and refresh failures.
GitOrigin-RevId: 756197d26f3fd347c28c400228f6b3d06ed493b7
## Why
MCP config reloads updated resolved server and authentication values, but did not
carry managed server constraints or plugin requirements into active threads.
## What changed
- Refresh MCP state from the typed thread config, including its managed
requirements, while retaining the thread's user layer and unrelated settings.
- Apply the refreshed config directly through `CodexThread` instead of encoding it
as a protocol operation.
- Re-evaluate effective plugins immediately after account changes so their MCP
requirements reach existing threads.
## Testing
- Cover replacement and enforcement of managed server and plugin requirements.
- Verify refreshed MCP settings are installed without changing the active model.
GitOrigin-RevId: 3cf0d69f51ef4893a4d722ee803fcbfdf1cd959e
## What changed
- Add `McpRefresh` to own the pending invalidation state and serialize MCP runtime refreshes.
- Keep cancellation recovery alongside the refresh state so a claimed refresh is marked pending again if publication does not complete.
- Update session initialization, shutdown, refresh paths, and tests to use the new abstraction.
GitOrigin-RevId: 915b28fa8b98a4f4c87ce8f2aa0d0ff99b57577d
## Why
Runtime updates reuse unchanged MCP connections, so an explicit
`RefreshMcpServers` operation could leave existing server connections intact.
## What changed
- Mark explicit MCP refreshes to rebuild connections for every configured
server on the next runtime replacement.
- Preserve the reconnect request if that replacement is cancelled before it
finishes.
- Continue reusing connections for ordinary runtime updates.
## Testing
Update the transient Apps startup recovery test to wait for both turn
completion and the MCP server's ready event before checking the recovered tool.
GitOrigin-RevId: cb9fd978b3c07543058a2393a154fb70911531de
## Why
Changes to MCP inputs can require rebuilding the runtime before the next model
step. Start that work eagerly so the step is less likely to wait for it.
## What changed
- Add a per-session worker that prewarms dirty MCP state after configuration,
authentication, or explicit refresh changes.
- Coalesce queued refreshes and always rebuild from the latest session state,
while keeping exact model-step refreshes as the correctness path.
- Stop and join the worker during session shutdown.
## Testing
- Verify rapid refreshes skip superseded configuration and publish the latest
state.
- Verify MCP-relevant policy changes mark the runtime for refresh.
GitOrigin-RevId: 045ed406caff97dcbb483bc183fd16eae0c7d7ad
## What changed
- Make `McpRuntime` own the published MCP configuration, connections, elicitation routing, and selected capability roots for a thread.
- Capture immutable MCP bindings for model steps and tool calls so in-flight work keeps a consistent connection set and approval authority while refreshed state is published atomically.
- Mark MCP state dirty when relevant configuration, plugins, environments, authentication, or elicitation capabilities change, then rebuild it before the next sampling request or out-of-band MCP operation.
- Separate config reloads from server invalidation: `ReloadMcpConfig` applies resolved MCP inputs, while `RefreshMcpServers` requests reinitialization from the thread's latest state.
## Testing
- Cover refreshed state visibility for existing turns, stable step bindings, current approval authority, cancelled refresh retries, resource-client reconciliation, and Apps recovery between sampling requests.
GitOrigin-RevId: 59eabb1aa8dc083426bd18ef4d3630508f376401
## Why
An idle thread with an outstanding durable sleep must resume when agent work
arrives, even when that message would normally remain queued without triggering a
turn.
## What changed
- Treat any pending mailbox message as wake-up work while a durable sleep is
attached to the thread.
- Keep requiring `trigger_turn` for idle threads that are not durably asleep.
- Cover waking a sleeping root thread from queue-only agent mail and persisting
the message in thread history.
GitOrigin-RevId: e361ac1b104436d4d63aebfe5e2422691a1cc139
## Why
Goal-first and forked threads could create rollouts without recording their
effective thread settings. Resuming those threads after a restart could then
use current configuration instead of the settings selected when the thread was
created.
## What changed
- Persist a `thread_settings_applied` item when a goal first materializes a
rollout, including a follow-up snapshot if settings changed concurrently.
- Append each fork's effective settings after its copied history so they take
precedence over inherited settings during resume.
- Centralize conversion from `ThreadConfigSnapshot` to
`ThreadSettingsSnapshot`.
## Testing
Added coverage that restarts the app server and verifies goal-first and forked
threads retain their respective `approvals_reviewer` settings.
GitOrigin-RevId: 6cb9aaae73bbe9ca432831cff7e826f95ffd6d01
## What changed
- Add the `SessionEnd` hook event to hook configuration, protocol schemas, analytics, and the hooks browser.
- Run matching hooks for root threads during shutdown, including app-server archive, delete, idle unload, and graceful shutdown. Flush the transcript first and provide the session ID, transcript path, working directory, and `reason: "other"`.
- Keep teardown bounded: hook output is advisory, the default timeout is one second, configured timeouts are capped at three seconds, and async hooks run synchronously with a warning.
## Testing
- Cover transcript availability, matcher selection, timeout normalization, ignored control output, subagent exclusion, archive/delete ordering, and graceful shutdown of multiple loaded threads.
GitOrigin-RevId: 5d19a658677a137caf836ed5042dcb43f5eb6d1a
## What changed
- Add a thread-owned `McpRuntime` that atomically publishes connection-set
replacements and owns shutdown.
- Route `McpResourceClient` snapshots and cache keys through that runtime so
extension clients observe refreshed MCP servers.
- Keep model-step configuration snapshots separate from ownership of live MCP
connections.
## Testing
- Add coverage that captures an extension's session resource client, refreshes
the MCP server configuration, and verifies that the existing client sees the
newly published server.
GitOrigin-RevId: 351b2e25594b8eb6b91e50ff8d06b7d49768ae1c
## Why
Multi-environment turns can run against different filesystems, so a single
thread-wide workspace-root list cannot accurately describe every selected
environment.
## What changed
- Add optional `runtimeWorkspaceRoots` to each app-server environment selection.
Omitted roots default to that environment's `cwd`, while an empty list selects
no roots.
- Carry environment-native roots through turn selection, model-visible context,
permission materialization, and tool sandboxing.
- Keep the top-level `runtimeWorkspaceRoots` field as a compatibility input when
constructing default environment selections; explicit selections use their
own roots.
## Testing
- Cover thread and turn API defaults, overrides, empty roots, and schema
round-tripping.
- Verify foreign roots remain model-visible and remote execution applies the
target environment's roots before sandbox selection.
GitOrigin-RevId: c6e7a62883c45c676c153055f46a55ee15a323a4
## Why
A thread resumed without an explicit reviewer could pick up the reviewer
from the current config instead of preserving the reviewer already in
use by the thread. After an app restart, this meant a thread running
with auto review could silently switch back to user review, and the next
turn could continue under the wrong reviewer.
## What changed
Persist the effective reviewer with each turn and restore the latest
persisted value when the thread resumes. If the resume request
explicitly provides a reviewer, that value still takes precedence.
## Test plan
- Added a regression test that starts a thread with auto review, records
a turn, restarts with user review in config, resumes without an
override, and verifies that auto review is preserved.
- `just test -p codex-protocol`
- `just test -p codex-state`
- `just test -p codex-rollout`
- `just test -p codex-app-server
thread_resume_preserves_persisted_approvals_reviewer`
- Clippy for the affected crates
## Why
A realtime session can end after more transcript has accumulated than
was included in its last handoff. That tail already lives in core's
active transcript state, but the stop path aborted the realtime
input/fanout tasks before routing it, so the final bit of the
conversation could disappear before `thread/realtime/closed`.
This behavior is still being evaluated, so clients must opt in per
realtime session. Omitted or false leaves shutdown behavior unchanged.
## What changed
- Add optional `flushTranscriptTailOnSessionEnd` to
`thread/realtime/start`, defaulting to false in app-server.
- Expose an idempotent `take_transcript_tail()` from the existing active
transcript state using `last_handoff_entry_count`.
- When enabled, let shutdown cancel the input owner cleanly and publish
at most one final existing `<realtime_delegation>` with the remaining
text in `<transcript_delta>`.
- Have the existing fanout drain already-parsed events before routing
that final delegation, so a queued handoff wins first and is not
duplicated in the tail.
- Flush realtime shutdown before ordinary session task abort during core
cleanup.
## Validation
- `just test -p codex-app-server-protocol`
- `just test -p codex-app-server
realtime_conversation_stop_emits_closed_notification`
- `just test -p codex-core conversation_transport_close_`
- `just test -p codex-core
conversation_close_routes_only_remaining_transcript_tail_once`
- scoped `just fix` for `codex-app-server-protocol`, `codex-app-server`,
`codex-protocol`, and `codex-core`
- `just fmt`
## Why
[#30867](https://github.com/openai/codex/pull/30867) makes
`submit_inter_agent_communication` the common outbound sink for
multi-agent v2 communications. This follow-up uses that single point to
log every communication lifecycle without requiring new hooks as spawn,
messaging, follow-up, or result paths evolve.
For each communication, the logs need to identify its type, sender and
receiver threads, and content, while correlating the successful send
with receipt by the destination mailbox. The logging path must not query
externally supplied time providers because those calls can be expensive
for app-server clients.
## What changed
- Added structured `INFO` events on the OpenTelemetry-exported
`codex_otel.agent_communication` target for `spawn`, `message`,
`followup`, and `result` communications.
- Logged successful sends from `submit_inter_agent_communication` with
the communication kind, sender and receiver thread IDs, content, and
submission ID.
- Logged receives after the communication has been enqueued in the
receiver mailbox, using the same submission ID.
- Avoided time-provider calls and other asynchronous work in the logging
path.
- Narrowed ordinary spawn and send-input APIs to `Vec<UserInput>` so
`Op::InterAgentCommunication` cannot bypass the context-bearing
centralized path.
The refactor does not change submission IDs, capacity checks, last-task
bookkeeping, mailbox ordering, protocol types, rollout data, or
model-visible context.
## Event shape
Illustrative JSON representation of the two independently emitted
records:
```json
[
{
"event.name": "codex.agent_communication",
"communication_id": "019f20e1-40d1-7890-a123-456789abcdef",
"kind": "spawn",
"state": "send",
"sender_thread_id": "019f20df-fbe1-7890-a123-456789abcdef",
"receiver_thread_id": "019f20e1-3f79-7890-a123-456789abcdef",
"content": "inspect the repository"
},
{
"event.name": "codex.agent_communication",
"communication_id": "019f20e1-40d1-7890-a123-456789abcdef",
"state": "receive"
}
]
```
Consumers join the receive record to the send record by
`communication_id` for the immutable communication metadata.
## Testing
- Extended the existing end-to-end multi-agent v2 spawn test to verify
content, both thread IDs, and a correlated send/receive submission ID.
- Re-ran focused control and handler coverage for direct messages,
follow-up tasks, and completion results.
### Summary
Release live thread persistence when a session ends because its
submission channel closes. This prevents a later same-process resume
from failing with `thread ... already has a live local writer`.
### Details
The issue is in the `codex-core` session teardown path used by Codex
hosts, rather than in Managed Agents API or exec-server itself.
Explicit shutdown already closes the `LiveThread`, which releases the
process-scoped writer held by `LocalThreadStore`. The
submission-channel-close fallback ran runtime and extension teardown but
skipped that persistence shutdown, leaving the thread ID registered as
having a live writer.
This change:
- closes the `LiveThread` on the channel-close fallback path;
- preserves the existing teardown order used by explicit shutdowns;
- extends the lifecycle regression test to assert that the thread store
receives `shutdown_thread`.
Context: [original
report](https://openai.slack.com/archives/C0B4NBHQGTV/p1782136364948039),
[recent occurrence
1](https://openai.slack.com/archives/C0B4NBHQGTV/p1782434817895839?thread_ts=1782136364.948039&cid=C0B4NBHQGTV),
[recent occurrence
2](https://openai.slack.com/archives/C0B4NBHQGTV/p1782335107474429?thread_ts=1782136364.948039&cid=C0B4NBHQGTV)
### Testing
- `just test -p codex-core
submission_loop_channel_close_runs_full_thread_teardown`
- `just test -p codex-core --lib` (1,989 passed; 3 skipped)
- `just fix -p codex-core`
- `just fmt`
- Native code review: no findings
I also attempted `just test -p codex-core`. The new regression passed;
79 unrelated integration tests failed in the local harness, primarily
because helper binaries such as `test_stdio_server` were unavailable,
plus local proxy/shell timing failures.
## Why
An MCP refresh can replace the session's current manager while a model
step is still running. The step must execute calls through the same
manager whose tools it advertised.
## Boundary
```text
current session MCP runtime
|
| capture once for this model step
v
StepContext.mcp
- exact MCP config
- exact connection manager
- exact runtime environment context
```
```rust
pub struct McpRuntimeSnapshot {
config: Arc<McpConfig>,
manager: Arc<McpConnectionManager>,
runtime_context: McpRuntimeContext,
}
```
## Example
```text
step A captures runtime A and advertises A's tools
refresh publishes runtime B
step A tool call -> runtime A
next step -> runtime B
```
Capturing the snapshot is only an `Arc` clone. It does not restart MCPs
or make an RPC.
## What changes
- Captures one MCP runtime in `StepContext`.
- Uses it for tool planning, tool calls, resources, approvals, connector
attribution, and elicitation.
- Publishes replacement runtimes atomically.
- Lets an old runtime live only while an in-flight step or request still
holds its `Arc`.
Most of this diff is mechanical routing from the session-global manager
to `step_context.mcp`; it does not introduce selected-plugin discovery
yet.
## What does not change
- No plugin or extension migration.
- No new MCP cache policy.
- No environment file watching.
- No client sharing between separate managers.
## Stack
1. Extension-owned World State sections.
2. Project executor skills through World State.
3. **This PR:** pin one MCP runtime to each model step.
4. Project selected MCP/app/connector metadata by environment
availability.
5. One end-to-end integration scenario.
## Why
Ultra should be one user-facing reasoning selection for work that
benefits from both maximum reasoning and proactive multi-agent
delegation. Without it, clients must coordinate maximum reasoning with
the experimental `multiAgentMode` setting, even though the inference
backend still expects its existing `max` effort value.
This change makes reasoning effort the source of truth: clients select
`ultra`, core derives proactive multi-agent behavior when the turn is
eligible for multi-agent V2, and inference requests continue to use the
backend-compatible `max` value.
## What changed
- Add `ultra` as a first-class reasoning effort and preserve
model-catalog ordering when exposing it to clients.
- Convert `ultra` to `max` at the inference request boundary, including
Responses HTTP/WebSocket requests, startup prewarm, compaction, and
memory summarization.
- Derive effective multi-agent mode per turn from effective reasoning
effort:
- eligible multi-agent V2 + `ultra` → `proactive`
- eligible multi-agent V2 + any other effort → `explicitRequestOnly`
- V1 or otherwise ineligible sessions → no multi-agent mode instruction
- Keep the derived effective mode in turn context history so successive
turns can emit a developer-message update only when the effective mode
changes.
- Remove selected multi-agent mode from core session configuration, turn
construction, thread settings, resume/fork restoration, and subagent
spawn plumbing. Subagents inherit reasoning effort and derive their own
effective mode.
- Retain the experimental app-server `multiAgentMode` fields for wire
compatibility while marking them deprecated. Request values are accepted
but ignored; compatibility response fields report `explicitRequestOnly`.
- Display Ultra in the TUI using the order supplied by `model/list`.
## Validation
- `just test -p codex-core ultra_reasoning_uses_max_for_requests`
- `just test -p codex-tui model_reasoning_selection_popup`
## Why
Once multi-agent mode can be selected per turn, clients also need to
choose the initial selection when creating a thread and observe that
selection through lifecycle and settings APIs.
The selected value is intentionally distinct from the effective
model-visible value: no client selection is represented as `null`, even
though an eligible multi-agent v2 turn derives `explicitRequestOnly` as
its effective default.
## What changed
- Add the optional experimental `thread/start.multiAgentMode` parameter
and pass it through thread creation.
- Preserve an omitted initial value as an unset selection rather than
eagerly storing `explicitRequestOnly`.
- Apply an explicit `thread/start` selection to the first turn through
the session configuration established at thread creation.
- Restore the latest persisted effective mode as the selected baseline
on cold resume when rollout history contains one.
- Inherit the optional selected mode from a loaded parent when creating
related runtime threads.
- Return the current selected `multiAgentMode` from `thread/start`,
`thread/resume`, `thread/fork`, and thread settings, using `null` when
no mode is selected.
- Keep lifecycle reporting independent from model capability and feature
eligibility; core turn construction remains responsible for calculating
and persisting the effective mode.
## Not covered
- Clearing an existing loaded-session selection back to unset through
`turn/start`; omitted or `null` currently retains the session's
selection.
- A TUI control, slash command, or `config.toml` preference.
## Verification
- `CARGO_INCREMENTAL=0 just test -p codex-app-server-protocol`
- `CARGO_INCREMENTAL=0 just test -p codex-app-server multi_agent_mode`
The focused app-server coverage verifies explicit `thread/start`
initialization, first-turn prompting, nullable reporting for an omitted
selection, and retention of selections that are not currently
runtime-eligible.
## Stack
Stacked on #28685. This PR contains only the thread initialization and
lifecycle/settings API layer.
## Why
Multi-agent v2 currently carries an explicit-request-only delegation
rule in its static usage hint. That provides a safe default, but it
prevents clients from selecting proactive delegation per turn without
changing static guidance or rewriting prior model context.
This change makes delegation mode a session selection that can be
updated through `turn/start`, while deriving the effective model-visible
mode separately for each turn. Eligible multi-agent v2 turns remain
explicit-request-only unless proactive mode is both selected and
enabled.
## What changed
- Add the experimental `turn/start.multiAgentMode` parameter with
`explicitRequestOnly` and `proactive` values. Omission retains the
loaded session's current optional selection.
- Add the default-off `features.multi_agent_mode` feature gate. Eligible
multi-agent v2 turns use the selected mode when enabled; an unset
selection or disabled gate resolves to `explicitRequestOnly`.
- Treat mode prompting as inapplicable for multi-agent v1 and other
unsupported session configurations, producing no multi-agent mode
developer message rather than rejecting the turn.
- Move the explicit-request-only rule out of the static v2 usage hint
and into a bounded, tagged developer context fragment.
- Emit the effective mode in initial context and only when that
effective mode changes on later turns.
- Persist the effective mode in `TurnContextItem` as the durable
baseline for resume and context-update comparisons.
Historical rollout items are not rewritten. Later mode developer
messages establish the current rule incrementally.
## Not covered
- Initial selection through `thread/start` and selected-mode reporting
from thread lifecycle/settings APIs; those are isolated in the stacked
#28792.
- A TUI control or slash command for selecting the mode.
- Persisting a preferred mode to `config.toml`; selection remains
session/turn scoped.
- Changes to multi-agent concurrency limits, tool availability, or model
catalog capability declarations.
- Rewriting historical rollout prompt items. Cold resume restores the
latest persisted effective mode when available while leaving historical
developer messages intact.
## Verification
- `CARGO_INCREMENTAL=0 just test -p codex-core multi_agent_mode`
- Focused app-server coverage verifies that `turn/start.multiAgentMode`
produces proactive developer instructions for an eligible v2 turn.
## Stack
Followed by #28792, which adds `thread/start` initialization and
lifecycle/settings observability.
## Stack
Depends on #28746. This PR implements shared rollout-budget accounting
and model-visible reminders using the configuration defined in #28746.
# Description / Main changes to Core:
`AgentControl` will now be the area where "rollout level" features &
accounting will have to live. It is incorrectly named for this
responsibility, but I think it can hold all the necessary shared state &
features (rollout token budget, mutliple thread interruption
responsibilitym etc)
In this PR, we have one "token ledger" that each thread will subtract
from when sampling. The "charge" will occur when response.completed() is
done and the calculation will be done on the responses api usage
carrier. The calculation will weigh sampling and pre-fill tokens as
specified.
Every time the budget crosses the configured reminder threshold, a
developer message is appended before the thread's next request
This remaining budget will _always_ be restated/reminded after a
compaction event.
Expiration and fan-out interruption will be in the stacked follow-up
(and also live in Agent Control).
## Reminders
"You have weighted {session_tokens_left} tokens left in the shared
session token budget."
The first request in each thread context receives the current remainder.
Later reminders are emitted after aggregate weighted usage crosses a
configured interval. If several intervals are crossed before a thread
sends another request, Core inserts one reminder with the latest
remainder.
Compaction response usage is charged before the next context starts. The
next reminder is appended after the compaction summary, leaving the
initial context content stable.
## Tests
Integration coverage verifies:
- weighted output and non-cached input accounting
- initial and periodic reminders
- shared accounting between a root and sub-agent
- post-compaction remainder and message placement
Local checks:
- `just fmt`
- `just test -p codex-core rollout_budget`
- `git diff --check`
The full workspace test suite was not run locally.
## Why
Realtime voice harness tuning needs app-side control over what backend
Codex text is spoken. Backend orchestrator text is written for a reading
UI, so automatically speaking every preamble, progress update, or final
assistant message can make the realtime voice model too chatty.
For experimentation, clients need two simple controls: keep app/client
text-item injection on the existing item-create path, and add an
explicit speakable path that app code can call only when it wants
realtime to speak. Automatic Codex output also needs an opt-in way to
switch from the protocol's default speakable path to regular realtime
items, with a caller-provided prefix so prompt wording can be tuned
outside core.
The default remains unchanged: if a client omits the new start fields
and never calls `appendSpeech`, automatic backend output continues down
the existing speakable path for the selected realtime protocol.
## What Changed
- Adds experimental `thread/realtime/appendSpeech` for app-provided
speakable text.
- Keeps existing `thread/realtime/appendText` as the item-create API for
app-provided realtime text items.
- Adds `codexResponsesAsItems` / `codex_responses_as_items` on
`thread/realtime/start` to send automatic Codex responses with
`conversation.item.create` instead of the protocol's default speakable
output path.
- Adds `codexResponseItemPrefix` / `codex_response_item_prefix` so
clients can prepend experiment instructions to those automatic Codex
response items.
- Keeps literal `conversation.handoff.append` routing scoped to the v1
speakable path; v2 default speech uses its item/function-output plus
`response.create` behavior.
- Removes the earlier public silent-context API and hardcoded
silent-context prefix.
- Updates realtime tests to cover default automatic speakable behavior,
opt-in automatic item-create behavior, and explicit `appendSpeech`
behavior.
## Validation
- `cargo check -p codex-core -p codex-app-server -p codex-api`
- `just test -p codex-app-server realtime_conversation`
- `just test -p codex-core realtime_conversation` (50/51 passed in the
filtered parallel run; the lone failure passed when rerun in isolation)
- `just test -p codex-core
conversation_mirrors_assistant_message_text_to_realtime_handoff`
- `just test -p codex-api
e2e_connect_and_exchange_events_against_mock_ws_server`
- `just fix -p codex-core`
- `just fix -p codex-app-server`
- `cargo build -p codex-cli`
## Why
Multi-agent v2 `send_message` deliveries already reach the receiving
model as typed `agent_message` items with encrypted content.
Child-completion notifications are generated by Codex itself, so their
content is plaintext and previously fell back to a serialized JSON
envelope inside an assistant message.
With plaintext `input_text` supported for `agent_message`, both delivery
paths can use the same model-visible type while preserving explicit
author and recipient metadata.
## What changed
- add plaintext `input_text` support to `AgentMessageInputContent` and
regenerate the affected app-server schemas
- preserve `InterAgentCommunication` as structured mailbox input instead
of converting it to assistant text
- record delivered communications as typed `agent_message` history items
- persist a dedicated rollout item so local delivery metadata such as
`trigger_turn` remains available without leaking into the Responses
request
- reconstruct typed agent messages on resume and preserve fork-turn
truncation behavior
- remove request-time assistant-content parsing
- preserve plaintext and encrypted inter-agent deliveries in stage-one
memory inputs
- normalize and link plaintext and encrypted agent messages in rollout
traces without treating inbound messages as child results
- cover the real MultiAgent V2 child-completion path end to end with
deterministic mailbox synchronization
## Verification
- `just test -p codex-core
plaintext_multi_agent_v2_completion_sends_agent_message`
- `just test -p codex-core input_queue_drains_mailbox_in_delivery_order
record_initial_history_reconstructs_typed_inter_agent_message
fork_turn_positions_use_inter_agent_delivery_metadata`
- `just test -p codex-memories-write
serializes_inter_agent_communications_for_memory`
- `just test -p codex-rollout-trace
agent_messages_preserve_routing_and_content
sub_agent_started_activity_creates_spawn_edge`
- `just test -p codex-rollout-trace
agent_result_edge_falls_back_to_child_thread_without_result_message`
- `just test -p codex-protocol -p codex-rollout -p
codex-app-server-protocol`
## Summary
We originally addressed startup prewarming holding the read side of
`RwLock<McpConnectionManager>` by snapshotting tool-list state. Review
feedback identified the broader ownership problem: the outer
synchronization should only publish or retrieve the current manager,
while MCP operations rely on the manager's internal synchronization. A
follow-up preserved operation retirement with a separate gate, but
further review questioned whether that synchronization was actually
required and whether we could support latest-wins replacement instead.
This PR now stores the current MCP manager in `ArcSwap`. Each operation
uses `load_full()` to obtain an owned `Arc<McpConnectionManager>`, then
performs MCP I/O without retaining the publication mechanism. Refresh
cancels obsolete startup work, constructs a replacement, and atomically
publishes it. New operations see the latest manager, while operations
that already loaded the previous manager retain a valid handle. Refresh
happens at a turn boundary, so there should be no active user tool calls
to drain.
Git history supports dropping the outer `RwLock`. It was introduced in
`03ffe4d595` on November 17, 2025 for non-blocking MCP startup: the
session published an empty manager, startup initialized that same object
while holding the write lock, and readers waited for initialization.
`7cd2e84026` on February 19, 2026 removed that two-phase initialization
in favor of constructing a fresh manager and swapping it in, explicitly
noting that `Option` or `OnceCell` could replace the placeholder design.
Hot reload later reused the existing lock to publish a replacement, but
I found no indication that the lock was introduced to guarantee
in-flight tool calls finish before refresh or shutdown.
Terminal shutdown remains separate from refresh: it aborts startup
prewarming and active tasks before shutting down the current manager, so
tool calls may be interrupted and no model WebSocket work continues
after shutdown. Focused regression coverage exercises pending tool-list
cancellation, deferred refresh, and startup-prewarm shutdown.
## Why
The realtime frontend model and the backing Codex thread should present
one coherent assistant. Raw typed messages, steers, and worker reports
belong to the orchestrator; the frontend model should receive the
orchestrator's user-facing result rather than a second copy of those
inputs.
Today normal `turn/start` input is automatically inserted into the
realtime conversation, while `turn/steer` is not. Besides creating
inconsistent context, this can make the frontend model react
independently before Codex has produced the response it should speak.
## What changed
- Remove automatic accepted-user-input mirroring into realtime
- Remove the mirror-only echo-suppression flag and dead V2 prefix helper
- Preserve explicit app-to-realtime text injection and FEM-to-Codex
delegation
- Replace the positive mirror tests and obsolete snapshots with a
negative routing regression test
## Test plan
- `cargo test -p codex-core
conversation_user_text_turn_is_not_sent_to_realtime`
- `cargo test -p codex-core
conversation_startup_context_is_truncated_and_sent_once_per_start`
- `cargo test -p codex-core inbound_handoff_request_starts_turn`
## Why
Thread cwd and environment selections are a single logical setting in
core: updating one without the other can silently desynchronize the
next-turn execution context. This change makes that relationship
explicit in the internal thread settings flow while preserving the
existing app-server public API shape.
## What changed
- Moved the cwd/environment pair through internal
`ThreadSettingsOverrides.environment_settings` instead of a top-level
internal `cwd` field.
- Kept `thread/settings/update` public params unchanged, with app-server
translating top-level `cwd` into the paired internal settings shape.
- Moved `Op::UserInput` environment overrides into thread settings so
user turns and settings updates use the same core path.
- Updated core, app-server, MCP, memories, sample, and test callsites to
construct the paired settings shape.
## Verification
- `git diff --check`
- Local test run starting after PR creation.
jif