## What changed
Update the bundled GPT-6-Astra instruction template to apply the
`functions.request_user_input_async` guidance only when the tool is available.
GitOrigin-RevId: 1dbf0299d4e5bf81e570e88c98b10c3822da99ed
## What changed
Update the GPT-6 instruction template to describe the available user-input tool directly, prohibit requests for file or screenshot uploads through its text-only interface, and allow 60 seconds for optional simple clarifications.
GitOrigin-RevId: acd186c58cee26f6dee20bd5de3a9775fe951231
## What changed
- Add optional per-model Guardian modes for computer use, shell, code mode,
file changes, MCP, network, and permission requests.
- Apply disabled, synchronous, or adaptive review by action category while
preserving legacy configuration when model policy is absent.
- Bind cached risk scores to the active model policy, thread settings, and
environment selections so outdated scores cannot approve later actions.
- Keep unknown review modes on the synchronous path and ignore unknown policy
fields for forward compatibility.
## Testing
- Cover policy serialization, legacy fallback, scoped tool classification,
disabled categories, and score invalidation after policy changes.
GitOrigin-RevId: f334a5f4fffb198da66cde3a400233e0df650516
## What changed
Describe the `gpt-6-astra` Fast service tier as providing 2x speed with
increased usage.
GitOrigin-RevId: a5b470dc598d23aea312172986823212922219f9
## What changed
- Add the hidden `gpt-6-astra` model definition, including its reasoning levels,
tool capabilities, context limits, agent instructions, and review policies.
- Reorder existing model priorities around the new catalog entry and provide the
browser-use policy to the bundled auto-review model.
- Keep TUI snapshots pinned to `gpt-5.6-sol` and make Guardian coverage read the
policy from the model selected by each test case.
- Allow the expanded `models.json` through the repository blob-size check.
GitOrigin-RevId: 039b224b91ca35c8d894cf6d23568bc6dc18ceeb
## What changed
- Add `auto_review.node_repl_policy` to model messages for `node_repl` and `cua_repl` reviews.
- Use the reviewer model's configured policy, falling back to the bundled policy when the field is absent and skipping injection when it is explicitly empty.
- Include the effective policy in Guardian session reuse checks and reject unsafe parent-model fallback transitions that would change it.
## Testing
- Cover catalog, bundled, and explicitly empty policies across Node and CUA review paths.
- Verify that policy changes invalidate cached reviewer sessions and incompatible fallback transitions.
GitOrigin-RevId: 51c0bdfed9b464091b6eb47d91ca5a0990d7a315
## What changed
- Add model metadata flags for enabling token budgeting and its history-notes extension.
- Apply those activation defaults once at thread startup when the user has not explicitly configured token budgeting, while respecting managed feature requirements.
- Continue resolving model-owned prompts for the active model so model switches can use their own token-budget guidance.
## Testing
- Update the token-budget integration test to verify that model metadata activates the feature without manual configuration.
GitOrigin-RevId: 73a32623eaeabb9243e240c43da7a5e7e93e9a19
## What changed
- Add model message metadata for built-in tool descriptions.
- Use the active step model's catalog description for `send_user_message_async`, including after a mid-turn model change.
- Fall back to the built-in description when catalog metadata is absent, while preserving an explicitly empty description.
## Testing
- Cover catalog serialization, fallback and empty-value behavior, and model changes within a turn.
GitOrigin-RevId: 72b953214ab9708931065321debf12133f8d2d40
## What changed
- Add an optional `proactive` multi-agent mode message to model metadata.
- Use the catalog's proactive message for `Ultra` reasoning when no general mode hint is configured. A missing value falls back to the built-in proactive instructions, while an empty value suppresses the mode message.
- Keep explicit mode behavior for other reasoning efforts and refresh the applicable catalog message when the model changes.
## Testing
- Cover proactive overrides, built-in fallback, empty-value suppression, mode-hint precedence, non-`Ultra` behavior, and model switches.
GitOrigin-RevId: da0a9ebd9b58cf04ef13a703c210e5da2eed0884
## What changed
- Allow `request_user_input` for optional questions that materially improve work quality.
- Continue with best judgment when the tool returns no answers.
- Keep permission requests, escalations, and other required input out of the tool; ask one concise plain-text question when progress cannot safely continue.
- Update the collaboration-mode preset test for the revised direct-question guidance.
GitOrigin-RevId: 8393621f0d8b59d8b164368ab7f2c8f4f6b3b14e
## What changed
- Add an optional `multi_agent_reasoning_effort` field to model metadata.
- When `Ultra` is selected, use a supported catalog override when present. Otherwise, prefer `max`, then the highest supported non-`ultra` effort, and finally `medium` when the model has no reasoning levels.
- Preserve existing request behavior for other reasoning efforts, including mapping `persistent` to `disabled`.
## Testing
- Cover configured, missing, unsupported, and empty-catalog fallbacks for both parent and spawned-agent requests.
GitOrigin-RevId: b7ef576f32e0ea95ef38bf95c25789ed16aae94e
## What changed
- Add optional Browser Use and Computer Use confirmation-policy Markdown to model catalog messages.
- Send the configured documents verbatim in `openai/confirmation_policies` metadata for `node_repl` and `cua_repl` tool calls. Send an empty object when the issuing model has no policies so runtime defaults are cleared.
- Use the issuing step's model snapshot across approval waits, and omit the metadata for other MCP servers and Guardian sessions.
## Testing
- Cover policy serialization, configuration overrides, model changes, approval waits, partial or empty policies, eligible server filtering, and Guardian exclusion.
GitOrigin-RevId: 038c4c262a31fc71986e4a2a947fe9bcbd407acc
## What changed
- Add bundled proactivity and follow-up guidance when `ReasoningEffort::Persistent` is selected.
- Allow model metadata to override or disable the guidance with `persistent_instructions`.
- Track the instructions as world state so effort and model changes replace or retire stale guidance without duplicating unchanged context.
- Tailor approval guidance to `send_user_message_async` availability, rebuild it for forked agents and rolled-back history, and exclude Guardian sessions.
## Testing
- Cover default, overridden, disabled, replaced, and retired instructions.
- Cover model changes, history transitions, agent forks, Guardian sessions, and asynchronous-message tool availability.
GitOrigin-RevId: 78e05d0335a70b385bd069b45edcd4ba570cef91
## What changed
- Add the experimental `amazonBedrockAccessKeys` app-server login flow, persist its credentials in the configured auth store, and use them for SigV4-signed Bedrock requests.
- Report `bedrockAccessKeys` as a distinct auth mode and document access-key login alongside AWS profile and environment credential setup.
- Clear the selected Bedrock provider, AWS settings, and Bedrock-specific model on logout while leaving external AWS credentials and `.env` untouched.
- Reject managed access-key login when higher-precedence configuration would prevent its region or profile settings from taking effect.
## Testing
- Cover access-key login, storage, request signing, configuration conflicts, status reporting, and logout cleanup.
GitOrigin-RevId: bbf297887c6503d3651ceb2aad85c4d626e32723
## What changed
- Use `exec_command` and `write_stdin` as the shell tool surface.
- Treat legacy `default`, `local`, and `shell_command` model metadata as `unified_exec`.
- Remove obsolete shell-selection configuration and runtime paths while preserving the feature and policy gates for zsh fork execution.
GitOrigin-RevId: d743cbe598630d73052f1fecad680c4cde17977d
## What changed
- Add the hidden Daybreak Blue and Daybreak Red model definitions.
- Refresh model capabilities, instructions, plan availability, and service-tier metadata.
- Configure the auto-review model for Responses Lite and code-mode tools, and update Guardian request tests for the resulting `additional_tools` and developer-message layout.
## Testing
- Update Guardian request snapshots and assertions for initial and follow-up reviews.
- Verify MCP approval routing against the developer-message form of the Guardian prompt.
GitOrigin-RevId: 6680e9abebdcbc43224a81348591e26f8422f3ec
## 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
## What changed
- Add `rejection_instructions` and `timeout_instructions` to catalog-provided auto-review messages.
- Use the acting model's instructions for denied and timed-out reviews across tool approvals, shell escalation, and MCP elicitation responses.
- Fall back to the existing instructions only when a catalog value is absent, while preserving explicit empty-string overrides.
## Testing
- Cover catalog overrides, legacy fallbacks, empty values, and separation between acting-model and reviewer-model messages.
GitOrigin-RevId: c5b2c2dbdaefd45d1d658651dd1abaeb6d8c93da
## What changed
- Allow `gpt-5.6-sol`, `gpt-5.6-terra`, and `gpt-5.6-luna` context-window overrides up to 872,000 tokens.
- Build the corresponding Amazon Bedrock entries from each bundled model's metadata while preserving Bedrock-specific identifiers, availability, tool, reasoning, and service-tier settings.
## Testing
- Cover long-context override clamping for bundled and Amazon Bedrock GPT-5.6 models.
- Verify the normalized Amazon Bedrock catalog metadata and context limits.
GitOrigin-RevId: 9f530dee501222a4d280bb31bbcdc2dacbe35d7d
## What changed
- Add optional `guardian_v2` model-message defaults for classifier instructions,
review thresholds, reasoning effort, transcript selection and limits, and
action and compaction token limits.
- Apply those defaults when Guardian v2 samples a tool call, while preserving
explicit `[features.guardianv2]` settings over catalog values.
- Preserve the new catalog configuration when applying model overrides.
## Testing
- Cover model-catalog serialization and model override behavior.
- Verify that sampling uses catalog defaults while retaining local overrides.
GitOrigin-RevId: 4ae9ec235c647acbca05a20b7c4b87a4e1331161
## What changed
- Add model-catalog messages for root and subagent roles, explicit delegation, and delegation hints.
- Resolve role instructions in config, catalog, then bundled-default order, while preserving empty values as an explicit way to suppress fallback text.
- Refresh catalog-provided role and mode instructions when the model changes, and give full-history forks the selected child model's subagent role without retaining the parent's role guidance.
- Keep existing config overrides and reasoning-effort behavior, including proactive delegation for ultra reasoning effort.
## Testing
- Cover message deserialization and preservation through model overrides.
- Cover precedence, empty overrides, model switches, resumed sessions, and full-history subagent forks.
GitOrigin-RevId: 4625cf7c6a5490176adddfaa0fb99100707daea9
## What changed
- Set `parallel_tool_calls` for regular and remote compaction prompts without consulting model metadata.
- Remove `supports_parallel_tool_calls` from `ModelInfo`, the bundled model catalog, and related fixtures.
- Preserve the existing Responses Lite behavior that disables parallel tool calls at request construction.
GitOrigin-RevId: 49552bdf97e71fa57325abb49f8adc0c88f1401a
## What changed
- Parse `node_repl_auto_review_required` and `node_repl_disabled` from model catalog entries and include their values in each turn's Responses API metadata.
- Treat both fields as reserved metadata so client-provided values cannot override the selected model's policy.
- Use the selected review model's catalog entry when building metadata for review turns.
## Testing
- Cover metadata propagation across model switches, review turns, Responses API requests, and MCP tool calls.
GitOrigin-RevId: c1f63426eeb8e559abff439fbc618716a4908aff
## What changed
- Add `include_apps_usage_instructions` to model metadata, defaulting to true for compatibility with existing metadata.
- Emit generic Apps guidance only when apps are available and the selected model enables it.
- Enable the capability for interactive model presets while leaving `codex-auto-review` and local models opted out.
## Testing
- Verify that missing metadata defaults to enabled and an explicit opt-out survives serialization.
GitOrigin-RevId: 6a20573cddfd3b80083d2b4b015829b49e6cc283
## What changed
- Propagate the optional `modelSpecialty` field through the model catalog and
app-server `model/list` response.
- When a newly selected model has the `cyber` specialty, default an active TUI
thread to workspace-write permissions with on-request approval. Use auto
review when available, otherwise keep the user as reviewer, while respecting
configured permission requirements.
- Show a notice when auto review is applied and strengthen the full-access
warning for cyber models.
- Preserve explicitly selected permissions when only reasoning settings change.
## Testing
- Cover auto-review defaults, requirement-aware fallback behavior, permission
preservation across reasoning changes, and the new warning notice.
GitOrigin-RevId: 8f25b0082d3051d9ff91f7355921236d0f82f758
## 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 `include_plugin_usage_instructions` to model metadata, defaulting to false.
- Emit generic plugin guidance only when plugins are available and the selected model enables it.
- Enable the capability for interactive model presets while leaving `codex-auto-review` opted out.
GitOrigin-RevId: 67f5a97e978033f5f1d533956c0b9deeae610283
## What changed
- Remove `ModelInfo.base_instructions` as an in-memory instruction source and use `model_messages.instructions_template` consistently for bundled, remote, fallback, and overridden model metadata.
- Preserve compatibility by promoting legacy `base_instructions` values when reading model responses and caches, and by including rendered legacy instructions when serializing `ModelsResponse` for older clients.
- Treat templates without instruction variables as literal text and retain the other model-message fields when applying instruction overrides.
This completes the consolidation proposed in https://github.com/openai/codex/pull/31302.
## Testing
- Cover legacy response and cache migration, canonical-template precedence, fallback instructions, personality rendering, overrides, and model switching.
GitOrigin-RevId: 089d986ca5e30da67db2c77a1b6a046d2cff52dc
## What changed
- Add optional `default` and `plan` collaboration-mode messages to model catalog metadata.
- Prefer the active mode's catalog message, including an explicit empty value, while falling back to the existing developer instructions when the catalog variant is absent.
- Track the model in collaboration-mode world state so mode and model changes append the appropriate instructions, and clear prior instructions when no current message exists.
- Preserve catalog collaboration messages when instruction overrides remove other model instruction fields.
## Testing
- Cover catalog deserialization, variant selection, fallback and empty-message behavior, legacy snapshot migration, mode changes, and model changes.
GitOrigin-RevId: d4c6c0028423bc50b0e8175e55586f3e7bc1c3a8
## Why
Matching model ETags can arrive on every response, but a recent models cache does
not need to be rewritten each time.
## What changed
Renew the models cache timestamp only after more than half of its configured TTL
has elapsed. Matching ETags continue to avoid refetching `/models`.
## Testing
Added integration coverage that verifies a recent cache remains unchanged and an
older cache is renewed without another `/models` request.
GitOrigin-RevId: b45da3824674c6b41d982ba4a405b419b09b7715
## What changed
- Add token-budget settings to model catalog messages and apply them when the
feature is enabled without explicit token-budget configuration.
- Keep explicit user settings authoritative, reject invalid catalog defaults,
and preserve resolved defaults in exported config locks.
- Manage context-window guidance through world state so it updates once when
the active model changes while retaining prior conversation history.
## Testing
- Cover catalog defaults, explicit overrides, disabled features, invalid
values, config-lock replay, and model switching.
GitOrigin-RevId: 54544fefaa14b09f5ef5bad9967a13c52b87b0fe
## What changed
- Add the action-risk review prompt template and default generic security policy to the `codex-auto-review` catalog entry.
- Define its `never` approval and `danger_full_access`, `workspace_write`, and `read_only` permission messages.
GitOrigin-RevId: 025bb609c12ae9d5235054747267a1ee46d118ef
## Why
ChatGPT token refresh requests need to follow the same configured routing policy as other authentication traffic, including environments that use the system proxy.
## What changed
- Require an `AuthRouteConfig` throughout login, logout, token refresh, personal access token, and agent identity flows.
- Build auth HTTP clients directly from that configuration instead of falling back to a default HTTP client when routing configuration is absent.
## Testing
- Add an integration test that refreshes a token through a cached system-proxy route and verifies the proxy receives the expected request.
GitOrigin-RevId: 7d54ab3219939a49921c51faf08cd4c2eabca51c
## What changed
- Add model-catalog approval message variants for `never` and `unless_trusted`.
- Select the catalog message that matches the active approval policy, while retaining the existing built-in text when that variant is absent.
- Treat an explicitly empty variant as an instruction to suppress the built-in approval text, consistent with `on_request` messages.
## Testing
- Cover variant selection, fallback and empty-message behavior, catalog deserialization, and the initial permissions message sent to the model.
GitOrigin-RevId: a0f8d41a08645f39b80093be53f200eeee18ca25
## What changed
- Add personality instruction variables for the GPT-5.6 model variants.
- Add the GPT-5.5 availability notice.
- Remove the `auto_review` and `permissions` message fields and the legacy `supports_reasoning_summary_parameter` flag from the bundled catalog.
GitOrigin-RevId: 426b0735a78232d2f658e0124db7c45889d08286
## What changed
- Set the bundled GPT-5.6 model variants' context windows to 272,000 tokens.
- Refresh model instructions and message configuration.
- Record reasoning-summary support and update skills, permissions, and auto-review metadata across the model catalog.
GitOrigin-RevId: 8be55124bf50af14756a174033a5cb362f4db714
## What changed
- Add an optional `policy_template` field to auto-review model messages.
- Build Guardian instructions from the catalog template when present, while
preserving the existing precedence for managed and catalog policy content.
- Fall back independently to the bundled template and bundled policy, and
preserve explicit empty catalog values.
## Testing
- Cover missing and empty template deserialization, policy precedence, bundled
fallbacks, and Guardian prewarm instructions.
GitOrigin-RevId: 93116dadbc9bf40527882844a428caed4159fff6
## What changed
- Hide `gpt-5.4` and `gpt-5.4-mini` from model selection and direct existing users to `gpt-5.6-terra` and `gpt-5.6-luna`, respectively.
- Use Terra for memory consolidation and Luna for memory extraction and rate-limit switch prompts.
- Keep Amazon Bedrock GPT-5 catalog entries visible when their bundled model metadata is hidden.
## Testing
- Add coverage that all Amazon Bedrock GPT-5 models remain visible and update TUI expectations for the new model choices.
GitOrigin-RevId: fd69c8d5fb6b510c77204df6b7ac3f07e67cded2
## 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
## Why
Embedding callers need to control whether model catalogs are persisted to disk.
## What changed
- Accept a shared models manager when constructing `ThreadManager` instead of always creating a cached manager internally.
- Add provider APIs for creating uncached model managers. OpenAI-compatible providers fetch on each `OnlineIfUncached` refresh without reading or writing `models_cache.json`.
- Keep existing app server, MCP server, sample, and test callers on the standard cached manager.
## Testing
- Verify an uncached manager fetches on every refresh.
- Verify an injected uncached manager controls thread refresh behavior and does not create `models_cache.json`.
GitOrigin-RevId: 8bb646054ecec4ccb865b2bf56249384916ea9db
## What changed
- Add an `auto_review.policy` field to model catalog messages.
- Use the selected Guardian model's catalog policy for review-session instructions, while preserving the precedence of `guardian_policy_config` and falling back to the built-in policy when neither is present.
- Preserve auto-review messages when model instruction overrides remove catalog instruction templates.
## Testing
- Cover configured-policy precedence, explicit empty catalog policies, catalog-message preservation, and propagation of the catalog policy into a prewarmed Guardian session.
GitOrigin-RevId: 26b61ae2958ea8325a64834dcf91f47e140d74b3
## What changed
- Add `supports_reasoning_summary_parameter` to model metadata, defaulting to
`true` for backward compatibility.
- Omit `reasoning.summary` and its summary-delivery stream option when the
selected model does not support the parameter.
- Apply the capability of the final selected model when a spawned agent uses a
different model.
## Testing
- Cover unsupported models in regular requests and spawned-agent model
overrides.
GitOrigin-RevId: 72b783799fc0685cef1501ef2dbf62d1308ceead
## Why
Model catalog instructions can include a baked-in `# Personality` section. An
explicit `none` setting should omit that section instead of sending it as part
of the model's base instructions.
## What changed
- Pass the configured personality into the models manager.
- When personality support is enabled and the setting is explicitly `none`,
remove the `# Personality` section through the next level-one heading from
catalog base instructions and instruction templates.
- Preserve explicit `base_instructions` overrides and avoid warning when no
personality was requested.
## Testing
Added unit and integration coverage for section removal, heading boundaries,
CRLF input, preserved configurations, and explicit base instructions.
GitOrigin-RevId: 452c88d3ac6001c2ac7d4fef269cd75dc239fa61
## What changed
- Build a reasoning payload for every Responses request and always include `reasoning.encrypted_content`.
- Remove `supports_reasoning_summaries` from model metadata and retire the `model_supports_reasoning_summaries` configuration override.
- Use configured or model-default reasoning effort without a capability gate, including for guardian reviews and tracing.
GitOrigin-RevId: 2c9f194a5d2d4d688a2235299e6358f82ab8e1ea
## Summary
External auth had two paths: provider-command credentials were resolved
through `ExternalAuth`, while app-provided ChatGPT credentials were
installed separately and only used the provider for refresh.
`ExternalAuth` also declared an auth mode independently from the
`CodexAuth` value it returned, so the declaration and credentials could
disagree.
This change makes the provider-owned `CodexAuth` authoritative for
initial resolution, credential kind, and refresh.
- remove `ExternalAuth::auth_mode` and require providers to return their
current auth from `resolve`
- route external auth registration, resolution, and unauthorized refresh
through `AuthManager::set_external_auth`
- keep the last resolved credential with its provider so synchronous
consumers and unauthorized recovery observe the credential's actual mode
- make the app-server bridge own both initial and refreshed ChatGPT
credentials, and detach it on logout
- keep model listing free of auth-refresh side effects, including in
offline mode
Provider-command auth still follows its configured cache interval.
App-provided ChatGPT auth still asks the parent app once after a `401`
and retries the request once.
Stacked on #31355.
## Testing
- `just test -p codex-login`
- `just test -p codex-models-manager`
- focused `codex-app-server` tests for external login/logout,
unauthorized refresh, and workspace mismatch
## 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
- 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`
rhan-oai