## What changed
- Carry each contextual fragment's `ContentItemKind` through initial context assembly and world-state updates.
- Attach the classifications to merged response messages through `content_item_kinds`, preserving their order alongside the message content.
- Convert extension prompt fragments into annotated developer fragments and classify built-in developer and guardian instructions.
## Testing
- Add an integration test covering roles and content annotations in the first request.
- Update world-state and persisted-history tests to account for annotation metadata.
GitOrigin-RevId: 71c48dad5b17bffd47ba03263e3eb6d5e74fbf4b
## What changed
Propagate tool cancellation tokens into Guardian approval reviews so interrupting a tool also aborts its pending review. Apply the same cancellation behavior to server-initiated MCP approval elicitation.
## Testing
Add integration coverage for cancellation from direct tool calls, code-mode turns, background code-mode cells, and MCP elicitation.
GitOrigin-RevId: cbee4b422bbcbf6984e415ef6409cee705beebdc
## Testing
- Verify an executor plugin's `Stop` hook starts running after its environment
attaches and stops after disconnection.
- Confirm hook calls carry the expected session, thread, turn, model, and request
metadata.
- Reject hooks whose MCP server belongs to a different executor environment.
- Cover the current restriction to the first executor environment and handler.
GitOrigin-RevId: ef13baf61379997f117cc57515363cc9880d3724
## Why
MultiAgent V2 workers can be reviewed after the root conversation has received
new user authorization that is absent from the worker's own transcript. Guardian
reviews need that genuine user context without treating forwarded or
assistant-authored claims as authorization.
## What changed
- Add the bounded root conversation to Guardian prompts for worker reviews.
- Preserve user and assistant roles, and explicitly treat only root user messages
as authorization evidence.
- Exclude summaries, review artifacts, and assistant commentary from the root
conversation evidence.
- Apply the same evidence to both core Guardian prompts and Guardian V2
classification.
## Testing
Add an integration test covering late root-user authorization, forged role text,
forwarded agent claims, and filtered synthetic authorization.
GitOrigin-RevId: 8846de69ba214a88f58f70675798d324e213427c
## What changed
- Remove the legacy `shell_command` handler and runtime, leaving `exec_command`
and `write_stdin` as the shell execution tools.
- Treat legacy `shell_command` model metadata as `unified_exec`, and normalize
legacy user opt-outs so they do not disable command execution. Managed feature
requirements and `shell_tool` can still disable it.
- Preserve shell approvals, sandboxing, zsh-fork support, and output truncation
through the unified execution path.
## Testing
- Cover legacy configuration and model-metadata compatibility.
- Exercise unified shell execution, approvals, truncation, and `apply_patch`
serialization across the app-server and core test suites.
GitOrigin-RevId: 5c2fd6164fc3519cdae4944cb9db276b8467311c
## Why
A checkout could point its `.git` file at a trusted repository's worktree
directory without proving that the repository had registered that checkout.
This could cause project configuration from an unrelated checkout to be treated
as trusted.
## What changed
- Verify the linked worktree's `gitdir` backlink, `commondir`, registered
checkout, and main checkout ownership before resolving the main repository's
trust key.
- Reject missing, oversized, symlinked, mismatched, or swapped Git metadata.
- Preserve valid linked worktrees that use path aliases, separate Git
directories, or non-UTF-8 POSIX paths.
## Testing
Add resolver and config-loading coverage for forged worktrees, metadata races,
case-sensitive paths, moved worktrees, and host MCP startup from project config.
GitOrigin-RevId: 6052a7d10ad2d613436f20175c356abdef8c758e
## Why
A model-provided shell path should select the requested shell type without
allowing that path to determine which executable Codex runs.
## What changed
- Resolve model-provided shells through Codex's normal shell discovery and
fallback logic after detecting their type.
- Keep the configured packaged zsh executable when the zsh-fork feature is
enabled and the file exists.
- Update shell, unified exec, and network approval expectations to use the
resolved local executable and arguments.
GitOrigin-RevId: ebe6f7eec2cfd1c0548d5bf1a26b7a30dba02cc2
## What changed
- Add `send_user_message_async` for root agents when the async-message feature is enabled and the model advertises tool support.
- Emit the supplied text as an asynchronous agent message and immediately return an accepted result so the turn can continue.
- Keep the user-visible update out of the model's input context.
## Testing
- Add an integration test covering tool registration, item events, continued turn execution, and model context.
GitOrigin-RevId: 13d5168cb8cd6742db17bb74a15c910994e9c98b
## Why
An unfamiliar executable can ignore its arguments, so trust in an apparent
inner command must not implicitly trust the executable that runs it.
## What changed
- Evaluate unfamiliar shell executables alongside their parsed commands when
determining unified exec approval requirements. Inner commands can add
restrictions, but cannot grant trust to the executable.
- Include the executable in reusable approval keys so approval for one custom
shell does not apply to another.
- Parse literal PowerShell scripts without launching the requested executable,
preserving command policy checks before approval.
## Testing
Add cross-platform coverage for spoofed shell paths, allowed and forbidden
inner commands, explicit custom-shell approval, and session approval isolation.
GitOrigin-RevId: 0dd2c7e9a2ac30965ef5fe5de1a8d2968bb5f9d0
## What changed
- Add a shared, read-only `ConversationHistorySnapshot` capability to the extension API.
- Include the snapshot in `ToolStartInput`, preserving conversation order while excluding contextual user messages.
- Avoid acquiring a history snapshot when no tool lifecycle contributors are registered.
## Testing
- Verify snapshots share existing response items, remain stable as history changes, and filter contextual user messages.
- Verify tool-start callbacks receive the history available for successive tool calls.
GitOrigin-RevId: 822cae33781dd876ea628f104be08d8a449a4900
## 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
## Why
Background refreshes previously warmed only the on-disk cache, so new sessions in the same process continued using the startup snapshot.
## What changed
- Make `CloudConfigBundleLoader` retrieve the latest shared bundle on each configuration load.
- Update the in-memory bundle after successful refreshes while preserving the last successful bundle when a refresh fails.
- Stop refresh work when its loader is replaced or its final clone is dropped.
## Testing
- Cover refreshed requirements and managed configuration in later sessions.
- Cover concurrent initial loads, refresh failures and recovery, bundle clearing, and refresh-task cancellation.
GitOrigin-RevId: ae3d79b575bad71a1eed370cee6dc66dc022ffa8
## What changed
- Filter saved `allow` prefix rules from the execution policy for cyber-specialized models and models listed in `auto_review.ignore_rules`, while preserving prompt, forbidden, network, and host-executable policy entries.
- Apply the filtered policy consistently to permission instructions, shell commands, unified exec, and the zsh-fork backend.
- Limit cyber-model command approvals to one-time decisions without proposing reusable policy amendments. General models continue to honor saved prefixes, including after switching models within a thread.
## Testing
- Add policy-level coverage for filtering allow-prefix rules while retaining restrictive and network rules.
- Add end-to-end coverage across shell and unified exec paths, zsh-fork execution, user and automatic review, and model switching.
GitOrigin-RevId: f3bd3872424291b12354ec415f33986ec369a368
## What changed
- Add the public asynchronous `ModelsCache` contract and cache entry/error types.
- Let model providers and `OpenAiModelsManager` accept caller-provided cache implementations while retaining the existing file-backed cache by default.
- Treat cache read failures as misses, keep cache write failures non-fatal, and support refreshing an entry's TTL after ETag revalidation.
- Continue to prefer an explicitly configured static model catalog over an injected cache.
## Testing
- Cover injected cache hits, read and write failures, TTL refreshes, and end-to-end model selection.
- Verify that the file-backed implementation satisfies the new cache contract and does not serve expired entries before revalidation.
GitOrigin-RevId: eef957a317910b8502a0c20f8fc112f689d5e389
## 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
## Why
MCP authentication checks need to use the same HTTP routing as the MCP
transport so servers reached through configured proxies can be discovered
reliably.
## What changed
- Resolve OAuth discovery and authentication status through each server's
runtime HTTP client for both local and managed environments.
- Keep local discovery capped at five seconds while allowing explicit login
requests to retain their requested timeout.
- Resolve refreshed MCP configuration and its runtime context from the same
snapshot.
## Testing
- Cover OAuth discovery through an environment proxy and macOS system proxy
resolution.
- Cover proxied MCP startup and runtime refresh with updated authorization
headers.
- Verify capped and preserved OAuth discovery timeout policies.
GitOrigin-RevId: 461fb1d4786e547df8b1e6b2215a8ac40438a3aa
## Why
Startup prewarms and Guardian review sessions do not need workspace Git metadata. Collecting it for those requests performs redundant work, while user turns still need a fresh view of the repository.
## What changed
- Add a per-session Git enrichment policy.
- Skip enrichment for startup prewarm, preview, and Guardian review turns.
- Keep fresh enrichment enabled for regular sessions and one-shot delegated turns.
## Testing
- Cover prewarm and Guardian requests omitting workspace metadata while subsequent user turns observe repository changes.
- Cover concurrent repositories and worktrees retaining distinct metadata.
- Verify concurrent enrichment is coalesced and canceled enrichment can be retried.
GitOrigin-RevId: 4a68615288bb568c75f2e487d9c0776e3e80f191
## Why
`codex-http-client` is the intended owner of direct `reqwest` usage.
## What changed
- Build Codex's default and route-aware clients with `HttpClientBuilder`, and return `HttpClient` directly to API, model discovery, auth, remote control, skills, and extension callers.
- Preserve custom CA, proxy, Cloudflare cookie, sandbox, and request-logging policies while removing `codex-login`'s direct `reqwest` dependency.
- Use shared HTTP types for auth headers, status codes, transport errors, and login-server tests.
## Testing
- Add a Linux integration test verifying that a regular Responses API turn traverses the configured system proxy.
GitOrigin-RevId: 714695fb80762d2083e2f7a6414183e2b3b63854
## Why
A fixed character limit does not account for the different context-window sizes supported by models.
## What changed
- Budget extension-rendered skill metadata at 2% of the resolved model context window, capped at 4,000 tokens.
- Keep the existing 8,000-character fallback when model context metadata is unavailable.
- Include the omission marker in the budget and still emit it when no skill entry fits.
- Apply the same resolved budget to executor and host skill catalogs assembled for a turn.
## Testing
- Cover proportional and capped budgets, multibyte fallback accounting, and omission-marker behavior.
- Verify through the production turn path that larger model context windows include more catalog entries without exceeding the computed budget.
GitOrigin-RevId: 4667293f1594de4dd605b32b9fa4255d26772c0f
## Why
Sandbox wrappers replaced the process launch command and did not carry an
`ExecParams.arg0` override through to the inner process.
## What changed
- Route sandboxed Unix launches with a custom `arg0` through a helper mode that
re-execs the requested program with the override.
- Expose the helper executable to the filesystem sandbox and dispatch its mode
from Codex and exec-server test binaries.
## Testing
Add coverage for the prepared sandbox command and an end-to-end remote process
that verifies both the custom `arg0` and filesystem restrictions.
GitOrigin-RevId: c9f8eef3906d184e670184c2eeed250d5895a9ca
## What changed
- Remove the `spawn_agents_on_csv` and `report_agent_job_result` tools and their agent-job runtime and state models.
- Drop the legacy `agent_jobs` and `agent_job_items` tables during state database migration.
- Keep `features.enable_fanout` and `agents.job_max_runtime_seconds` accepted as no-op compatibility settings while omitting them from the generated configuration schema.
## Testing
- Verify upgrades remove both legacy agent-job tables.
- Verify the removed feature and configuration keys still parse without taking effect.
GitOrigin-RevId: 8cc3337da78c67162229f02f40a747f503542646
## Why
Audio attachments could be dropped from tool outputs and reconstructed thread history, while inline base64 payload size could overstate their model context cost.
## What changed
- Preserve remote and local audio attachments in legacy user-message events and thread history.
- Keep audio from function and custom tool outputs when the model supports it and the output fits the truncation budget; otherwise replace it with an omission marker.
- Estimate inline audio cost from its duration, with a serialized-size fallback for malformed or unsupported data URLs, so compaction and truncation use model-visible cost instead of raw base64 size.
- Replace audio in messages and tool outputs with an explanatory placeholder for models without audio input support.
## Testing
Add coverage for history round trips, dynamic tool responses, modality filtering, duration-based context estimates, output-budget truncation, and remote compaction.
GitOrigin-RevId: 5ed4628850a668f8383b7cd73db71b144cc02cd1
## What changed
- Add an optional `initialItems` field to `thread/realtime/start` for seeding
realtime V3 session history with `user`, `developer`, and `assistant` text.
- Encode the items in Frameless Bidi session bootstrap payloads for websocket
and WebRTC transports while preserving existing payloads when the list is
omitted or empty.
- Reject non-empty initial items for V1 and V2, more than 128 items, any item
over 8,192 estimated tokens, or a combined total over 8,192 estimated tokens.
## Testing
- Cover protocol serialization, app-server forwarding, websocket and WebRTC
payloads, empty-list compatibility, version checks, and size limits.
GitOrigin-RevId: 09af367f9a07543e3f8d59b7fa5387f932dc5a6a
## What changed
Add a Unix integration test that starts two `exec-server` environments with
separate workspace roots. Verify that each environment can write within its own
workspace while cross-workspace writes are denied.
GitOrigin-RevId: 98f2f58b97c1dcca8e103702b067c65d4eeed0c4
## Why
Starting a new session should not have to wait for an unchanged stdio MCP
server to initialize before its tools can be presented to the model.
## What changed
- Cache recent tool catalogs for matching stdio MCP server configurations and
use them while a new live connection starts in the background.
- Bound the process-wide cache to 32 entries with a 30-minute TTL, and key
entries by the server configuration, environment, working directory, and
elicitation capabilities that affect the catalog.
- Keep connection-specific instructions and tool annotations out of cached
catalogs. Resolve tool calls against the live connection so tools that are
unavailable or hidden in the new session cannot be invoked from stale data.
- Publish only the newest completed catalog fetch and do not cache HTTP servers
or stdio configurations that depend on remotely sourced environment values.
## Testing
- Add coverage for cache expiration, environment changes, catalog sanitization,
and out-of-order refreshes.
- Add an end-to-end test that starts inference from cached definitions while a
replacement MCP process initializes, then verifies calls and visibility
against the live process.
GitOrigin-RevId: 2ed0f29ba53ace12532730f951b175c588aa3ae7
## What changed
- Add per-sandbox-mode permission messages to `ModelMessages` and preserve them when applying model configuration overrides.
- Use the selected catalog message in permission instructions, substituting `{{ network_access }}` with the active network policy. Fall back to the existing sandbox text when no override is provided, and allow an empty override to omit only the sandbox section.
- Apply catalog permission messages when a session starts and when its model changes.
## Testing
- Cover catalog deserialization, mode selection, network substitution, fallback and empty-message behavior, remote model catalogs, and model changes.
GitOrigin-RevId: cd5ed3aee3155dca0e7b2358c0f09bae73236856
## What changed
- Default `prompt_cache_key` to the session ID from response metadata instead of the thread ID.
- Preserve explicit prompt cache key overrides.
- Verify that API-key-authenticated root and subagent requests use the same session-based cache key even though they have different thread IDs.
GitOrigin-RevId: cb1bf21a3cb8fa2c5083c5c6afd4829ff2adcf1f
## Why
After a cold resume, a Multi-Agent V2 root thread must still be able to
address its persisted descendants even though their runtimes are not reopened
eagerly.
## What changed
- Restore metadata for open descendant agents when resuming a V2 root thread.
- Preserve each descendant's path, nickname, and role while leaving its runtime
unloaded until a message targets it.
- Reuse the spawn metadata preparation path so restored identities retain the
same reservation and naming checks as newly spawned agents.
## Testing
Added an end-to-end cold-resume test that follows up with a persisted agent by
name and verifies that the original agent thread is loaded on demand.
GitOrigin-RevId: a1663c1ae60eefadb948cd39f933cd3bbb49a492
## What changed
- Stop the TUI and app server from inspecting existing sessions at startup to
set `personality = "pragmatic"` automatically.
- Remove the migration marker, helper APIs, and associated tests.
GitOrigin-RevId: 9e963bf113aff21651696bbc463f2749ac0f510a
## What changed
- Pass configured workspace roots from core to the exec server so filesystem and process sandbox permissions are materialized against the intended roots.
- Preserve an explicitly empty workspace-root list instead of treating the sandbox working directory as an implicit root.
- Initialize filesystem sandbox contexts with their working directory as the default workspace root.
## Testing
- Add end-to-end coverage for patch and command writes inside and outside workspace roots.
- Verify remote filesystem and process sandboxes do not grant access through an empty workspace-root list.
GitOrigin-RevId: 6684c8f7de50970b45a29e2a6323df954c96f9f6
[Codex Thread
019f2408-dc59-79f2-b245-4c11debd1a61](https://codex-thread-link.openai.chatgpt-team.site/thread/019f2408-dc59-79f2-b245-4c11debd1a61)
## Why
Long-lived Codex sessions can outlive the ChatGPT bearer token that was
present when the MCP runtime started.
The Responses path already recovers from token expiration by refreshing
or reloading the shared `AuthManager`. The reserved `codex_apps`
hosted-plugin client did not observe that update: `McpConnectionManager`
built its `/ps/mcp` HTTP auth once from a `CodexAuth` snapshot, and
`auth_provider_from_auth` copied that snapshot bearer into a static
`BearerAuthProvider`.
After the copied bearer expired, `/ps/mcp` kept sending it even though
Responses had a newer token in the same `AuthManager`. The failure
occurred before downstream connector execution, so unrelated apps such
as Gmail, Slack, and Google Calendar could all fail with the same
transport-level `401 token_expired`.
This replaces
[openai/codex#29474](https://github.com/openai/codex/pull/29474), which
was closed for inactivity without being merged. A new long-lived-session
report reproduced the same simultaneous `/ps/mcp` expiry pattern across
unrelated apps.
## What changed
- Add an `AuthManager`-backed request-header provider in
`codex-model-provider`. It keeps an `Arc<AuthManager>` and reads
`auth_cached()` for each outbound request, so the next `/ps/mcp` call
sees a token refreshed by the existing Responses/auth-recovery flow.
- Scope that provider to the startup account, ChatGPT user, and
workspace identity. Same-identity token reloads are followed; an account
switch emits no ambient auth until account-scoped MCP state is rebuilt.
- Have `McpConnectionManager` construct the dynamic provider only for
the reserved `codex_apps` registration used by the hosted-plugin
`/ps/mcp` path.
| MCP path | Auth behavior after this change |
| --- | --- |
| Reserved `codex_apps` hosted-plugin `/ps/mcp` | Read current
same-identity auth from the shared `AuthManager` per request |
| `codex_apps` with `CODEX_CONNECTORS_TOKEN` | Keep the environment
bearer-token override |
| User-configured/direct MCP registrations | Keep their existing
configured auth path |
## Non-goals
- No plugin-service changes.
- No downstream Slack, Gmail, Calendar, or other connector
OAuth/link-refresh changes.
- No auth UI changes.
- No behavior change for user-configured/direct MCP registrations.
- No new `/ps/mcp`-initiated token refresh; this makes `/ps/mcp` observe
refreshes already performed through the shared `AuthManager`.
## Tests
- `just test -p codex-model-provider`
- Covers same-identity token reloads and refuses a changed startup
identity.
- `just test -p codex-mcp`
- `just test -p codex-core mcp_auth_refresh`
- Creates the reserved hosted-plugin `codex_apps` `/ps/mcp` client
before the shared `AuthManager` changes, updates that same manager
through its public external-auth path, performs a real `tools/call`, and
asserts the request uses the current bearer.
### summary
We want to pause code-mode from yielding back to the model when a
subcommand triggers an approval prompt. This means that all of these
previously inline blocking requests should also take out a
ElicitationService registration.
This also does some plumbing refactoring to request patch approval to
make it match the other `request_*_approval` methods in that it blocks
on the approval in the function instead of returning the oneshot
channel, this affords our ability to encapsulate the ElicitationService
registration via RAII.
Adds tests to confirm the blocking behavior for code_mode both in suite
tests and that the session holds them.
## Summary
- enable `auth_elicitation` by default
- promote the feature to `Stable`, as default-enabled features must be
stable
- update the feature regression test to cover the new lifecycle state
and default
## Impact
Auth elicitation is now available without requiring users or clients to
opt in through configuration.
## Testing
- `just test -p codex-features` (52 passed)
## Summary
- complete unified-exec processes from the ordered event stream instead
of issuing a final zero-wait `process/read`
- add optional executor sandbox-denial state to `process/exited`
- retain `process/read` as a retained-output and compatibility fallback
for receiver lag, sequence gaps, and legacy servers
- recover sandbox-denial state across transport reconnection
- cover the real `TestCodex` remote-exec path without adding a public
test-only event constructor
## Why
A successful one-shot tool call currently receives its output and
terminal notifications, then pays another wide-area `process/read` round
trip before returning. Staging traces showed that remote response wait
accounted for more than 99.8% of RPC time; local serialization,
queueing, and deserialization were below 0.6 ms.
## Measured impact
A direct staging A/B used the same build and route and changed only
completion mode. Each arm ran three times with 30 one-shot
`/usr/bin/true` calls per run. The table reports the median of the three
per-run percentiles.
| Metric | Final `process/read` | Pushed events | Change |
| --- | ---: | ---: | ---: |
| End-to-end completion p50 | 159.5 ms | 118.7 ms | -40.8 ms (-25.6%) |
| End-to-end completion p95 | 182.4 ms | 131.7 ms | -50.6 ms (-27.8%) |
| Completion-wait p50 | 80.1 ms | 41.5 ms | -38.5 ms (-48.1%) |
| Final `process/read` RPC p50 | 79.9 ms | eliminated | -79.9 ms |
TCP_NODELAY was enabled in both A/B arms, so its effect cancels out. The
successful, complete, in-order event path issued zero final
`process/read` calls.
## Compatibility and recovery
- new servers send `sandboxDenied` on `process/exited`
- legacy servers omit it, which triggers one compatibility
`process/read`
- broadcast lag or a sequence gap triggers a retained-output read
- recovery remains bounded by the server's existing 1 MiB
retained-output window
- complete, in-order event streams issue no completion read
- sandbox denial is attached to the exit event before consumers can
observe process completion
- server-first and client-first rollouts remain wire-compatible;
server-first realizes the latency win immediately
## Integration coverage
The `TestCodex` suite exercises four distinct remote-exec contracts:
- complete pushed output/exit/close with zero reads
- direct pushed sandbox denial with zero reads
- legacy missing denial metadata with exactly one compatibility read
- count-bounded replay eviction recovered from retained output without
duplication
## Validation
- `just test -p codex-core
exec_command_consumes_pushed_remote_process_events`: 4 passed
- `just test -p codex-core unified_exec::process_tests::`: 4 passed
- `just test -p codex-exec-server`: 294 passed, 2 skipped
- `just test -p codex-exec-server-protocol`: 5 passed
- `just test -p codex-rmcp-client`: 89 passed, 2 skipped
- focused Bazel `//codex-rs/core:core-all-test`: passed across 16 shards
- scoped `just fix` passed for core and exec-server
- `just fmt` passed
The complete workspace suite was not rerun; focused Cargo and Bazel
coverage passed for the changed behavior.
## Summary
MCP refresh replaced the published connection manager without shutting
down the manager it superseded. If another task retained that old
manager, its stdio MCP processes stayed alive and accumulated across
refreshes.
Atomically swap in the refreshed manager, then explicitly shut down the
exact manager returned by the swap. Add a process-level regression test
that retains the old manager during refresh and verifies its stdio
process exits while the replacement remains available.
## Context
Explicit cleanup was lost when manager publication moved to `ArcSwap`.
Dropping the old manager is not a reliable shutdown boundary because
active callers can retain its `Arc` and underlying client process
handles.
Responses API safety buffering metadata currently stops at the transport
boundary, so app-server clients cannot render the in-progress safety
review state.
This change:
- decodes and deduplicates `safety_buffering` metadata from Responses
API SSE and WebSocket events without suppressing the original response
event
- emits a typed core event containing the requested model plus backend
use cases and reasons
- forwards that event as `turn/safetyBuffering/updated` through
app-server v2 and updates generated protocol schemas
- keeps the side-channel event out of persisted rollouts and turn timing
This supports the Codex Apps buffering UX and depends on the Responses
API backend work in https://github.com/openai/openai/pull/1044569 and
https://github.com/openai/openai/pull/1044571.
Validation:
- focused `codex-core` safety-buffering integration test passes
- `cargo check -p codex-core -p codex-app-server -p
codex-app-server-protocol`
- `just fix -p codex-api -p codex-protocol -p codex-core -p
codex-app-server-protocol -p codex-app-server -p codex-rollout -p
codex-rollout-trace -p codex-otel`
- `just fmt`
- broad package test run: 4,430/4,492 passed; 62 unrelated
local-environment/concurrency failures involved unavailable test
binaries, MCP subprocess setup, and app-server timeouts
Stacked on #28766.
## Why
Network approvals are environment-scoped: allowing a host in one
execution environment should not allow the same host in another
environment.
#28766 adds the inert IDs and constructor plumbing. This PR applies the
behavior on top.
## What changed
- Route managed network traffic through per-environment HTTP and SOCKS
proxy listeners.
- Stamp HTTP, HTTPS CONNECT, SOCKS TCP, and SOCKS UDP policy requests
with the source environment at the proxy boundary.
- Carry the selected execution environment through shell, unified exec,
zsh-fork, and sandbox transform paths.
- Include the environment in pending, approved-for-session, and
denied-for-session network approval cache keys.
- Include the environment in approval IDs and approval prompts.
- Preserve legacy fallback for unattributed requests, but deny when
active-call attribution is ambiguous.
- Fail closed if an environment-specific proxy endpoint cannot be
prepared.
## Validation
- just fmt
- CI will run tests and clippy
## 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.
## Why
`child_agents_md` is a disabled, under-development experiment that adds
a second model-visible explanation of hierarchical `AGENTS.md` behavior.
Keeping it leaves unused prompt, configuration, documentation, and test
surface.
## What changed
- remove the `ChildAgentsMd` feature and `child_agents_md` config schema
entry
- remove the hierarchical prompt asset, export, and instruction
injection
- remove feature-specific tests and documentation
- keep the generic unstable-feature warning coverage using
`apply_patch_streaming_events`
Normal project `AGENTS.md` discovery and composition are unchanged.
## Testing
- `just test -p codex-features`
- `just test -p codex-prompts`
- `just test -p codex-core agents_md`
- `just test -p codex-core unstable_features_warning`
Stacked on #28822.
## Summary
- add a host-injectable current-time provider with a built-in system
implementation
- record UTC developer reminders in history immediately before due model
requests
- keep cadence state per session and force a refresh after compaction
This does NOT include the app server client <-> server clock logic. This
PR is only for the reminder message & system clock that will be used in
prod.
## Testing
- `just test -p codex-core varlatency_`
- `just clippy -p codex-core -p codex-app-server -p codex-mcp-server -p
codex-thread-manager-sample`
- `just fmt`
## 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
Som tools, such as history and notes, must remain top-level when MCP
deferral is enabled while staying unavailable through code-mode `exec`.
## What changed
- Added `features.code_mode.direct_only_tool_namespaces`.
- Classified matching MCP tools as `DirectModelOnly`.
- Kept those tools top-level in `code_mode_only`.
- Excluded them from `tool_search` deferral and the nested `exec`
surface.
- Updated the generated config schema.
## Validation
- `code_mode_only_exposes_direct_model_only_mcp_namespaces`
- `load_config_resolves_code_mode_config`
## Why
Image generation used `std::fs::read` for referenced image paths, which
did not support environment-backed filesystems or their sandbox context.
## What changed
- Expose optional turn environments to extension tool calls.
- Include each environment’s ID, working directory, filesystem, and
sandbox context.
- Read referenced images through the selected environment filesystem.
- Keep sandbox usage at the extension call site so extensions can choose
the appropriate access mode.
- Consolidate image request construction into one async function.
- Add coverage for successful environment reads and read failures.
## Validation
- `cargo check -p codex-image-generation-extension --tests`
- `just fmt`
- `just bazel-lock-update`
- `just bazel-lock-check`
`just test -p codex-image-generation-extension` could not complete
because the build exhausted available disk space.
## Why
The model should be able to see bounded context-window budget metadata
when the `token_budget` feature is enabled. The full-window message is
only injected with full context, while normal turns get a smaller
follow-up only when reported usage first crosses a budget threshold.
## What changed
- Added the `TokenBudget` feature flag.
- Added `<token_budget>` developer fragments for full context-window
metadata and current-window remaining tokens.
- Inserted the threshold message during normal turn handling by
comparing token usage before and after sampling, avoiding persistent
threshold bookkeeping.
- Added core integration coverage for full-context-only metadata and
25/50/75 percent threshold messages.
## Verification
- `just test -p codex-core token_budget`
- `git diff --check`
## Why
Multi-Agent V2 concurrency should count active non-root turns, not
resident or durable agent threads. The limit is intentionally best
effort: admission checks are synchronous, but concurrent successful
checks may overshoot slightly.
## What changed
- Keep one root-derived execution limit on the shared `AgentControl`.
- Count active V2 subagent turns with an RAII guard owned by
`RunningTask`.
- Check capacity before spawning or starting an idle agent, including
direct app-server `turn/start` submissions.
- Preserve queued delivery for agents that are already running.
- Exempt automatic idle continuations so `/goal` work is not dropped
when capacity is temporarily full.
- Keep root and V1 turns outside this limiter.
## Test coverage
- `execution_guards_count_active_v2_subagent_turns`
- `execution_guards_ignore_root_and_v1_turns`
- `v2_nested_spawn_checks_shared_active_execution_capacity`
## Why
This PR fixes approval sandbox semantics in the unified-exec path. The
zsh-fork runtime exposed the bug because the shell can do meaningful
work before any intercepted child `execv(2)` exists: redirections,
builtins, globbing, and pipeline setup all happen in the launch process.
If the model requested `sandbox_permissions=require_escalated`, or an
exec-policy `allow` rule explicitly bypassed the sandbox, that approved
sandbox decision needs to be preserved for the launch path and for
intercepted execs that use the same approval machinery.
The behavior is not only about zsh fork. The production changes are in
shared approval/escalation code, so they also affect non-zsh-fork
intercepted exec paths that go through the same sandbox decision logic.
The narrow intent is to preserve the approval decision while still
keeping denied-read profiles and bounded additional-permission requests
sandboxed.
## Production Changes
- `codex-rs/core/src/tools/runtimes/unified_exec.rs`: derives a
`launch_sandbox_permissions` value from the requested sandbox
permissions and the runtime filesystem policy, then uses that value for
managed-network/env setup and launch sandbox selection. This keeps full
approval or policy-bypass decisions visible to the first unified-exec
attempt, while still preventing a full sandbox override from discarding
denied-read restrictions. Direct unified exec keeps the same decision
surface; the important difference is that zsh-fork launch setup no
longer accidentally loses the approved parent sandbox decision.
- `codex-rs/core/src/tools/runtimes/shell/unix_escalation.rs`: makes
intercepted-exec escalation selection explicit for the three sandbox
permission modes. `UseDefault` only escalates when an exec-policy
decision allows sandbox bypass, `RequireEscalated` escalates when
unsandboxed execution is allowed, and `WithAdditionalPermissions`
escalates through the bounded additional-permissions path instead of
being treated as a full unsandboxed override. Unsandboxed intercepted
execs now also rebuild the environment as `RequireEscalated`, which
strips managed-network proxy variables consistently with other
unsandboxed execution.
## Test Coverage
Most of the PR is tests. The new coverage verifies:
- unified exec preserves parent approval and exec-policy sandbox
decisions for zsh-fork launch selection;
- bounded `with_additional_permissions` remains sandboxed and
permission-profile based;
- denied-read profiles are not weakened by parent approval;
- explicit prompt rules still prompt for intercepted execs after the
parent command is approved;
- unsandboxed intercepted execs strip managed-network env vars.
No documentation update is needed; this is an internal approval/sandbox
correctness fix.
---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed
with [ReviewStack](https://reviewstack.dev/openai/codex/pull/24981).
* #24982
* __->__ #24981
## Summary
Responses Lite does not execute hosted Responses tools, so models using
it must route web search and image generation through Codex-owned
executors & standalone Response's API endpoints.
This PR is stacked on #26487.
## Validation
- `cargo test -p codex-core responses_lite_ --lib`
- `cargo test -p codex-core
standalone_executors_remain_hidden_without_flags_or_responses_lite
--lib`
- `cargo test -p codex-core
hosted_tools_follow_provider_auth_model_and_config_gates --lib`
- `cargo test -p codex-web-search-extension -p
codex-image-generation-extension`
- `cargo test -p codex-app-server --test all standalone_`
- `cargo fmt --all -- --check`
## Why
Guardian auto-review normally uses the provider-preferred review model
when one is available. Some parent models need model-catalog metadata to
select a different review model while keeping older `/models` payloads
compatible when that metadata is absent.
## What changed
- Added optional `ModelInfo::auto_review_model_override` metadata to the
public model payload as a review-model slug.
- Updated Guardian review model selection to prefer the catalog override
when present, while preserving the existing provider preferred-model
path and parent-model fallback when it is omitted.
- Added focused Guardian coverage for override and no-override model
selection.
- Added an `auto_review` core integration suite test that loads override
metadata from a remote model catalog path and asserts the strict
auto-review `/responses` request uses the catalog-selected review model.
- Updated existing `ModelInfo` fixtures and local catalog constructors
for the new optional field.
## Validation
- `cargo test -p codex-protocol
model_info_defaults_availability_nux_to_none_when_omitted`
- `cargo test -p codex-core guardian_review_uses_`
- `cargo test -p codex-core
remote_model_override_uses_catalog_model_for_strict_auto_review --test
all`
- `just fix -p codex-protocol`
- `just fix -p codex-core`
- `just fmt`
- `git diff --check`
pakrym-oai