## Why
#31335 lets HTTP callers obtain proxy-aware clients from
`HttpClientFactory`, but a non-HTTP transport such as WebSockets also
needs two pieces of policy owned by `codex-http-client`: a concrete
route decision for its destination and the same custom-CA-aware rustls
trust configuration used by HTTPS.
Keeping these prerequisites in the shared abstraction means the
dependent Responses WebSocket change (#31441) cannot independently
reinterpret `features.respect_system_proxy`, PAC results, or enterprise
CA settings.
## What changed
- Add a redaction-safe `OutboundProxyRoute` with explicit
transport-default, direct, and concrete-proxy outcomes.
- Add `HttpClientFactory::resolve_proxy_route()` so transports can
resolve a destination through the already-selected outbound proxy
policy.
- Resolve `ws://` and `wss://` URLs through their HTTP equivalents so
system and PAC rules apply consistently.
- Add an always-returned rustls config builder that starts from native
roots and layers in any configured Codex custom CA bundle. The existing
optional builder remains available to callers that can delegate the
default configuration to their transport.
- Continue redacting proxy URLs from `Debug` output because they may
contain credentials.
## Review guide
1. `http-client/src/outbound_proxy.rs` defines the transport-neutral
route result and WebSocket URL normalization.
2. `http-client/src/custom_ca.rs` factors the native-root/custom-CA
construction so callers that perform TLS themselves can always obtain a
config.
3. `http-client/src/outbound_proxy_tests.rs` verifies WebSocket
normalization and legacy transport-default behavior.
## Test plan
- `just test -p codex-http-client outbound_proxy`
- `just test -p codex-http-client custom_ca`
---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed
with [ReviewStack](https://reviewstack.dev/openai/codex/pull/31342).
* #31431
* #31363
* #31362
* #31361
* #31442
* #31441
* __->__ #31342
## Why
Test callers need one composable way to create app-server fixtures
instead of a growing family of overlapping constructor implementations.
## What
- add a feature-complete TestAppServer::builder()
- make the default builder own a temporary CODEX_HOME and select the
automatic test environment
- expose builder knobs for no automatic environment, explicit
CODEX_HOME, program, arguments, plugin startup tasks, environment
overrides, managed config, and JSON logging
- keep the existing public constructor surface, but route every
constructor through the builder so the new path is exercised immediately
- remove the redundant private constructor ladders; caller migration and
public constructor removal live in the optional cleanup stack
## Validation
- just test -p codex-app-server (940/941 before updating the expected
builder error wording)
- just test -p codex-app-server
auto_env_rejects_explicit_environment_config
- just fix -p codex-app-server
- just fmt
## Follow-up stacks
Cleanup, optional for the benchmark work:
1. [#31451 test: migrate TestAppServer callers to
builder](https://github.com/openai/codex/pull/31451)
2. [#31452 test: remove TestAppServer
constructors](https://github.com/openai/codex/pull/31452)
Benchmark infrastructure:
1. [#31427 test: add delayed exec-server
transport](https://github.com/openai/codex/pull/31427)
2. [#31295 bench: add cold skill load
macrobenchmark](https://github.com/openai/codex/pull/31295)
3. [#31428 bench: add e2e benchmark
entrypoints](https://github.com/openai/codex/pull/31428)
4. [#31429 ci: smoke Bazel e2e
benchmarks](https://github.com/openai/codex/pull/31429)
## Summary
Autocomplete completion previously inserted a new space even when a
separator was already present, which could leave redundant whitespace
around neighboring text. Popup dismissal also tracked only the query
string, so dismissing one token could suppress a different occurrence
with the same text.
This gives completions one horizontal-separator policy across files,
images, skills, and mentions. Existing separators are reused when
possible, ordinary suffix text remains separated, and line breaks stay
intact. Dismissal now identifies the complete whitespace-delimited token
occurrence, so offset-only edits preserve dismissal while later
identical tokens remain independent. Newly completed values beginning
with `$` or `@` can also remain closed when the next PR's affinity rule
recognizes the token to the left of the cursor.
## Examples
The examples below use `|` to represent the cursor.
### Reuse existing separators
Starting with an existing two-space gap:
```text
@ma| next
```
After accepting `src/main.rs` and typing `foo`, completion previously
added a third separator and left both original spaces before `next`:
```text
src/main.rs foo| next
```
After this PR, completion reuses the first existing separator as the
insertion point and preserves the second between the new text and
`next`:
```text
src/main.rs foo| next
```
### Keep a completed sigil-prefixed value closed
Starting before a line break:
```text
@ma|
next
```
and accepting a path whose result is itself prefixed with `@` produces:
```text
@scope/main.rs |
next
```
Without this PR's completion dismissal, the affinity rule in #30463
rediscovers the completed token to the left of the cursor and reopens
its popup:
```text
@scope/main.rs |
^^^^^^^^^^^^^^^ popup reopens
next
```
With this PR, the inserted occurrence is dismissed and the popup remains
closed:
```text
@scope/main.rs |
^^^^^^^^^^^^^^^ popup remains closed
next
```
### Do not dismiss an identical later occurrence
Given two identical tokens:
```text
@scope/main.rs| @scope/main.rs
```
after dismissing the first popup with Escape and moving to the second
token, query-only dismissal previously suppressed the second popup too:
```text
@scope/main.rs @scope/main.rs|
^^^^^^^^^^^^^ popup remains closed
```
After this PR, dismissal also matches the token's ordinal among complete
tokens, so the second occurrence opens normally:
```text
@scope/main.rs @scope/main.rs|
^^^^^^^^^^^^^ popup opens
```
### Keep dismissal across offset-only edits
After dismissing `@ma`, moving to its start, and pasting an email-like
token:
```text
email@ma.com @ma|
```
the `@ma` bytes embedded in `email@ma.com` do not count as another
autocomplete token. The original `@ma` keeps its dismissal even though
its byte range moved.
## Stack
This is PR 2 of 3, stacked on #31190. It relies on the explicit
replacement ranges introduced there and provides the completion
lifecycle used by the targeting fix in #30463.
## Summary
- recognize embedded HTTP(S) URL prefixes case-insensitively in Windows
dangerous-command detection
- add regression coverage for uppercase and mixed-case schemes inside
`Start-Process` invocations
## Why
PowerShell and URL parsing treat schemes case-insensitively, but the
pre-parser only searched for lowercase `http://` and `https://`. When a
URL appeared in the same shlex token as surrounding PowerShell syntax,
such as `Start-Process('HTTPS://example.com');`, the prefix was not
stripped and the command was incorrectly classified as not dangerous.
Validated with the scoped `codex-shell-command` suite (138 tests) and a
direct classifier reproduction that failed before the change and passed
afterward.
## Why
Pre-sampling compaction intentionally uses the previous turn's model
when the compaction compatibility hash changes or when switching to a
model with a smaller context window. This keeps compaction aligned with
the settings that produced the history, but it can block the next turn
when a resumed ChatGPT thread still references a model slug that has
since been retired. The Codex backend rejects that compaction request
before the user's currently selected model gets a chance to sample.
This PR lets those threads recover without changing previous-model
compaction behavior for API-key authentication or custom providers. It
is stacked on #31316, which is a behavior-preserving extraction of the
individual remote compaction attempts; this PR contains the fallback
behavior.
## What changed
- For automatic previous-model compaction, capture the selected model's
request context when using ChatGPT authentication with the OpenAI
provider and the selected model differs from the previous model.
- If the previous-model attempt returns an `InvalidRequest`, retry
compaction once with the selected model for both `/responses/compact`
and Responses Compaction V2.
- Complete history processing, lifecycle events, and token accounting
with the context of the model that successfully compacted the thread.
- If the fallback also fails, return the original previous-model error
so the retry does not change the user-visible failure.
- Record fallback attempts with reason, implementation, and outcome
telemetry.
- Leave API-key authentication, custom providers, same-model turns, and
non-`InvalidRequest` failures on their existing paths.
## Testing
- `just test -p codex-core -E 'test(pre_sampling_compact) |
test(model_unavailable_error)'` (10 tests)
- Added integration coverage for a resumed thread whose model was
renamed, a model downshift using Responses Compaction V2, and API-key
authentication with a custom provider.
This PR depends on [#31296](https://github.com/openai/codex/pull/31296)
for the canonical-to-legacy event mappings.
## Description
This PR makes the v1 and v2 wait paths emit canonical
`TurnItem::CollabAgentToolCall` lifecycle instead of
`CollabWaitingBegin` / `CollabWaitingEnd` directly.
Both paths already used the same legacy waiting events before this PR.
The v1 item carries receiver metadata and final agent statuses for its
target agents; v2 waits for mailbox activity rather than specific
agents, so it keeps those fields empty, matching the existing v2 legacy
payload.
App-server v2 consumes the canonical item directly and ignores the
mapped legacy wait events.
## Why
Wait is separate from the other collab tools because it is multi-target
and has distinct timeout/status behavior. Keeping it last also lets this
PR remove the old helper that only existed to shape legacy wait status
entries in core.
## What changed
- Emit canonical collab wait items from both v1 and v2 wait handlers.
- Preserve receiver metadata and agent status snapshots on completed
wait items.
- Remove the old core helper for building legacy wait status entries.
## Follow-up
The next stack PR, [#30188](https://github.com/openai/codex/pull/30188),
writes canonical `TurnItem` values to paginated rollout files.
This PR depends on [#31296](https://github.com/openai/codex/pull/31296)
for the canonical-to-legacy event mappings.
## Description
This PR makes the non-wait v1 collaboration tools—spawn, send input,
resume, and close—emit canonical `TurnItem::CollabAgentToolCall`
lifecycle instead of their legacy begin/end events directly.
App-server v2 consumes the canonical collab items directly, ignores the
mapped legacy events, and applies close-agent thread-watch cleanup from
the completed item.
## Why
These four tools share the same single-target lifecycle shape. Wait
stays separate because it carries multi-target status snapshots and has
its own status-shaping cleanup.
## What changed
- Add shared helpers for emitting canonical collab tool-call lifecycle.
- Migrate spawn, send input, resume, and close handlers.
- Move close-agent watcher cleanup onto canonical completed collab
items.
## Why
The backend config-bundle contract now exposes managed configuration in
`managed_layers`, split into `baseline` and `system_overlay`. The Rust
transport models need to match that contract before any runtime behavior
changes.
## What changed
- add the generated `DeliveredManagedLayers` model
- model `baseline` and `system_overlay` as required arrays
- expose optional/null `managed_layers` on delivered config and
requirements documents
- retain `enterprise_managed` for transport compatibility
This PR changes transport types only; cloud-config runtime behavior is
unchanged.
## Stack
1 of 4. Next: #31286.
## Validation
- `just test -p codex-backend-openapi-models -p codex-backend-client`
- `just test -p codex-cloud-config`
- revalidated against the current generated backend schema
This PR depends on [#31296](https://github.com/openai/codex/pull/31296)
for the canonical-to-legacy event mappings.
## Description
This PR makes the MultiAgentV2 spawn, message/follow-up, and interrupt
paths emit completed canonical `TurnItem::SubAgentActivity` items
instead of `SubAgentActivityEvent` directly.
App-server v2 now applies interrupted-agent thread-watch cleanup from
the canonical completed item and ignores the mapped legacy activity
event.
## Why
Sub-agent activity is separate from the v1 collab tool begin/end
lifecycle. Keeping it separate makes the v2 watcher side effect
reviewable without mixing in the larger collab tool-call migration.
## What changed
- Emit canonical sub-agent activity items from v2 spawn,
message/follow-up, and interrupt paths.
- Move missing-thread watcher cleanup onto canonical completed activity
items.
- Update focused app-server coverage to exercise canonical interrupted
activity.
## Why
This PR is a behavior-preserving refactor only. It does not add a
fallback, change which model is used for compaction, or otherwise change
compaction behavior. The behavioral change is implemented in the stacked
follow-up, #30319.
Pre-sampling compaction deliberately uses the previous turn's context
when the compaction compatibility hash changes or when switching to a
model with a smaller context window. That preserves the model settings
that produced the history being compacted, but the previous context is
not always usable. For example, a resumed thread can still reference a
model slug that has since been retired, causing compaction to fail
before the currently selected model can sample.
#30319 addresses that failure mode by retrying compaction with the
current turn's selected model when the backend rejects the
previous-model attempt. This PR performs only that preparatory refactor.
## What changed
- Extracted one legacy `/responses/compact` request attempt into
`compact_remote_request.rs`.
- Extracted one Responses-based remote compaction request attempt into
`compact_remote_v2_attempt.rs`.
- Kept hooks, lifecycle events, analytics, window advancement, history
processing and installation, and error behavior unchanged in the
existing orchestration paths.
- Preserved standalone Responses-based compaction's owned client-session
lifetime through lifecycle completion.
## Testing
- `just test -p codex-core -E 'test(remote_compact)'` (22 tests)
## What
Adds an optional hosted login-success redirect path for app-server login
requests.
- Keeps the existing localhost success page as the default.
- Lets app-server callers opt into a hosted success page with an
optional protocol field.
- Persists credentials before redirecting to the hosted success page.
- Keeps org setup and existing CLI/device-code login flows on the local
success page.
- Accepts an optional typed `appBrand` value and forwards it to the
hosted page as `app_brand` so web can select the correct asset.
- Generates the app-server protocol schema updates for the new optional
fields.
## Why
This supports the hosted Codex login success page rollout without
changing existing login behavior by default. The Codex Apps frontend can
gate the opt-in with Statsig after the hosted web page.
## Rollout safety
- Old callers omit the new field and continue using localhost.
- New callers talking to old app-server builds remain safe because the
Codex Apps side treats the field as optional and defaults the flag off.
- Missing brand values default to Codex.
- The hosted redirect always uses the app-login source so the hosted
page can reopen Codex; the existing streamlined-login visual flag
remains separate.
## Validation
- `just fmt`
- `just fix -p codex-login -p codex-app-server-protocol -p
codex-app-server -p codex-app-server-test-client -p codex-tui`
- `just test -p codex-login`
- `just test -p codex-app-server-protocol`
- `just write-app-server-schema`
- `git diff HEAD --check`
The focused login and protocol run passed all 380 tests.
I also started the broader `just test -p codex-app-server`; it compiled
successfully, then many tests failed on this machine because spawned
test servers tried to use the ambient `/Users/rafaelj/.codex/sqlite`
state DB, which is read-only in this sandbox. I stopped that run after
confirming the failures shared that environment issue.
## 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)
## Why
Approval guidance is currently assembled entirely by the client. Model
Messages V2 needs model catalogs to provide model-specific `on_request`
guidance for both user-reviewed and auto-reviewed approval flows while
retaining the existing generated prompt as a compatibility fallback.
## What changed
- add nullable `on_request` and `on_request_auto_review` catalog
messages
- select the message matching the active approvals reviewer for
`on_request` policies
- replace the complete legacy approval section when the selected catalog
value exists, including support for an empty string that suppresses the
section
- retain legacy rendering when the object or selected key is absent, and
for non-`on_request` policies
- preserve approval messages when base-instruction or personality
overrides clear instruction templates
- refresh permissions instructions when the active model changes
- pass catalog messages through initial and incremental permissions
construction
## Relationship to reviewer persistence
PR #31309 independently persists the approvals reviewer in turn context
and refreshes permissions when that reviewer changes. This PR is based
directly on `main` and does not duplicate that rollout migration; once
both land, reviewer switches will also select and append the new catalog
variant.
## Testing
- `just test -p codex-protocol`
- `just test -p codex-prompts`
- `just test -p codex-models-manager`
- `just test -p codex-core permissions_messages`
## Summary
- measure successful legacy `app/list` latency with
`codex.apps.installed.duration_ms`, segmented by `path=legacy` and
`reload`
- measure successful host-owned `codex_apps` startup and explicit
refresh latency with `codex.apps.refresh.duration_ms`
- add the refresh trigger to successful
`codex.mcp.tools.fetch_uncached.duration_ms` samples for `codex_apps`
without changing other MCP-server samples
## Why
This establishes a small latency baseline for the current connector path
before `ConnectorRuntimeManager`, `app/installed`, and `app/read` land.
Error-rate and broader runtime-state instrumentation are intentionally
deferred.
This is telemetry-only and does not change connector behavior.
## Validation
- `just test -p codex-mcp` (94 passed)
- `just test -p codex-app-server list_apps` (13 passed)
- `just fix -p codex-mcp`
- `just fix -p codex-app-server`
- `just fmt`
- `git diff --check`
## Summary
- Treat streamed Responses `bio_policy` failures as terminal invalid
requests instead of retryable stream errors.
- Recognize the new biology policy code and message in the TUI while
preserving the legacy `invalid_prompt` contract.
- Keep the existing dedicated biology safety notice and add
regression/snapshot coverage for all supported error shapes.
## Why
[openai/openai#1068559](https://github.com/openai/openai/pull/1068559)
gates a Responses API contract change from `invalid_prompt` and the
legacy message to `bio_policy` and new biology copy.
Without this compatibility change, streamed blocks are retried as
transient failures and the OSS TUI falls back to a generic/raw error
instead of the dedicated safety notice.
## Validation
- `just test -p codex-api` — 137 passed
- `just test -p codex-tui
app_server_safety_access_errors_render_dedicated_notice` — passed
- `just fix -p codex-api`
- `just fix -p codex-tui`
- `just fmt`
- `just test -p codex-tui` — 2,957 passed; two reproducible failures
remain in untouched Guardian feature-flag persistence tests:
-
`update_feature_flags_disabling_guardian_clears_review_policy_and_restores_default`
-
`update_feature_flags_disabling_guardian_clears_manual_review_policy_without_history`
## Summary
- Adds `writes` to `AppToolApproval` and exposes it through config and
app-server schemas, including
`[apps._default].default_tools_approval_mode`.
- In `writes`, tools with `readOnlyHint = true` skip approval; all other
tools prompt, including non-destructive writes and tools without
annotations.
- Prevents session or persistent approval choices in this mode so later
writes still prompt.
## Why
`auto` only prompts for risk-hinted actions, while `prompt` also
interrupts reads. Apps need a middle mode that gates writes without
prompting for declared read-only actions.
## Validation
- `just write-config-schema`
- `just write-app-server-schema`
- `just fmt`
- `just test -p codex-core mcp_turn_metadata` (4 passed)
- `just test -p codex-core writes_mode` (2 passed)
- `just test -p codex-app-server config_read_includes_apps` (1 passed)
- `just test -p codex-app-server-protocol` (251 passed)
- `just test -p codex-config` (200 passed)
- `just test -p codex-cli` (300 passed)
- `just fix -p codex-core -p codex-config -p codex-app-server-protocol
-p codex-app-server -p codex-cli`
## Why
Windows Bazel test shards currently cap local test actions at 4.
Controlled forced-test-execution measurements in #31339 found that 8
local jobs reduced Bazel elapsed time on every shard by 19–29%,
including the slow shard from 835.5s to 609.3s, while 12 jobs regressed.
## What
Set `common:ci-windows-cross --local_test_jobs=8` in `.bazelrc`.
## Manual validation
- `git diff --check`
- `just fmt`
- [8-job CI
experiment](https://github.com/openai/codex/actions/runs/28838387129)
## Summary
- refactor `ExternalAuth` to return `CodexAuth` directly
- remove the parallel `ExternalAuthTokens` wrapper
- preserve existing external bearer and app-server refresh behavior
This is a mechanical precursor refactor; it does not add auth
capabilities or change recovery behavior.
## Testing
- `cargo fmt --all`
- `cargo test -q -p codex-login --lib`
- `cargo test -q -p codex-app-server --test all
external_auth_refreshes_on_unauthorized`
## Why
Intel macOS release binaries crash on the first Code Mode tool call
while V8 creates its code range. The x86_64 V8 allocator later makes a
non-`MAP_JIT` reservation executable, which Hardened Runtime rejects
when the signature contains only `com.apple.security.cs.allow-jit`.
Tracks
[SE-8006](https://linear.app/openai/issue/SE-8006/intel-macos-codex-cli-crashes-in-v8-startup-on-gpt-56-sol-tool-calls).
Fixes#28390.
## What
- add an expanded entitlement profile only for x86_64 `codex` and
`codex-app-server`, the release binaries that link V8
- keep arm64 and `codex-responses-api-proxy` on the existing narrower
profile
- share one fail-closed target/binary selector between signing and final
verification
- verify the expected Mach-O architecture and exact entitlement
dictionary for the signed binary, tar.gz, zstd, package, and DMG copies
## Verification
- `just test-github-scripts` (34 tests)
- `UV_CACHE_DIR=/private/tmp/codex-uv-cache just fmt-check`
- `bash -n .github/scripts/macos-signing/select_codex_entitlements.sh`
- `plutil -lint` on both entitlement profiles
- parsed `rust-release.yml` as YAML
- `git diff --check`
- ad-hoc Hardened Runtime signing smoke on an x86_64 Mach-O slice:
strict `codesign` verification passed; the Codex profile contained
exactly both keys and the proxy profile retained exactly `allow-jit`
## Release validation
Run a native Intel smoke of the final Developer ID-signed x86_64 Codex
binary through V8 isolate creation before shipping. PR #30849 is
diagnostic scaffolding, but its non-sandbox release job currently fails
in the harness before V8 starts, so it is not counted as coverage here.
## Why
Before changing plugin namespace loading performance, lock down the
existing behavior so the same cases can be validated before and after
the optimization.
## What
- cover mixed plain and nested plugin skills under one scan root
- cover inherited, nested, and invalid-manifest namespace precedence
- cover symlinked plugin skill directories, symlinked plain directories,
and scan-root ancestor symlinks
## Validation
- just test -p codex-core-skills namespace on unmodified main: 12 passed
This PR depends on [#31296](https://github.com/openai/codex/pull/31296)
for the canonical-to-legacy event mappings.
## Description
This PR makes dynamic tools emit canonical `TurnItem::DynamicToolCall`
lifecycle instead of `DynamicToolCallRequest` /
`DynamicToolCallResponse` directly.
App-server v2 now sends the client `DynamicToolCall` request from the
canonical item start. It ignores the mapped legacy request/response
events, so clients receive one item start and one tool request.
## Why
Dynamic tools are a separate migration boundary because their start
event also drives a client request. Keeping that routing change isolated
makes it easier to verify that the request still happens exactly once.
## What changed
- Emit in-progress and completed/failed dynamic tool items from the
dynamic tool handler.
- Move app-server client request dispatch onto canonical dynamic item
starts.
- Add focused app-server coverage for the canonical start notification
and client request.
Michael Bolin