Adapt the Agent Plugin runtime boundary enforcement to the alpha.6.4 APIs and skip symlinks while installing plugins.
(cherry picked from commit 56b82e676c)
(cherry picked from commit 720c9d68e1)
## Why
Amazon Bedrock supports hosted text web search, but it rejects the
`search_content_types` field used for multimodal search and does not support
external live or indexed web access.
## What changed
- Advertise hosted web search for Amazon Bedrock while marking external web
access as unsupported.
- Resolve unsupported live and indexed modes to cached search, or disable the
tool when cached search is prohibited by managed requirements.
- Normalize built-in and configured Bedrock model catalogs to text-only web
search, and retain the runtime provider in session configuration so turn
setup can apply its capabilities.
## Testing
- Cover cached fallback, managed-mode restrictions, text-only tool payloads,
provider capabilities, and catalog normalization.
GitOrigin-RevId: 310473849257401654388a4ebb42920e03aa3228
## What changed
- Keep complete MCP namespace descriptions in tool-search source metadata.
- Raise the namespace tool-spec description limit from 1,000 bytes to 512 KiB,
truncating at a UTF-8 character boundary only when the new limit is exceeded.
## Testing
- Cover descriptions beyond the former limit and multibyte truncation at 512 KiB.
- Update SSE and stdio MCP tests to verify that complete server instructions are
preserved without hiding tools.
GitOrigin-RevId: 000bfcafb3df348065ae451685bfbf978a0e3248
## What changed
- Write staged `rusty_v8` checksum manifests with LF line endings on every platform.
- Strip carriage returns when verifying manifests so existing Windows-built releases with CRLF line endings remain usable.
## Testing
- Assert that staged checksum manifests contain no carriage returns.
GitOrigin-RevId: 142855c147a08a60f5e5782c61e2fcd6b2e350d6
## What changed
- Add `parse_agent_plugin_mcp_config` to translate Agent Plugins v1 `mcp.json` files into Codex MCP server configuration.
- Normalize `stdio` and streamable HTTP transports, including `PLUGIN_ROOT` and `PLUGIN_DATA` expansion, contained plugin paths, secure endpoint validation, and filtering of client-owned HTTP headers.
- Keep valid sibling servers when another server is invalid, while returning per-server parse errors.
- Preserve UTF-8 HTTP header values when forwarding streamable HTTP protocol headers.
## Testing
- Cover transport mapping, placeholder expansion, path containment, schema and field validation, per-server errors, platform-specific environment handling, and UTF-8 headers.
GitOrigin-RevId: ed4ab0fcf495afbb381ce48beb93989629444c56
## Why
Timing out a Git metadata command must not leave helper processes running after
the command wrapper exits.
## What changed
- Run Git metadata commands in a dedicated process group on Unix and a Job
Object on Windows so timeout cleanup terminates their full process trees.
- Start Windows commands suspended, assign them to the Job Object, and then
resume them so immediate descendants cannot escape containment.
- Preserve descendants when a Git command completes normally, and retain the
existing direct-spawn fallback if Windows Job Object setup fails.
## Testing
Added cross-platform regression tests for cleanup both while the command wrapper
is running and after it exits, plus Windows coverage for immediate-child Job
Object containment.
GitOrigin-RevId: 351851708e23ff06b89fe1894bd09a3558f67293
## 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
## Testing
- Verify consecutive tool reviews reuse the same Guardian session without
prewarming and execute both approved commands.
- Verify interrupting a pending Guardian review prevents command execution and
records an aborted tool result for the next turn.
GitOrigin-RevId: b8371ae82b4fef206857012589668f890f9305bf
## Why
MCP tools can be exposed directly, discovered through tool search, or called
from Code Mode. Servers need to be able to opt out of any of these surfaces
without disabling their tools everywhere.
## What changed
- Add `omit_tools_from` to MCP server configuration, accepting any combination
of `direct`, `deferred`, and `code_mode`.
- Apply the exclusions independently when building direct, deferred, and Code
Mode tool surfaces while keeping omitted tools registered for permitted uses.
- Remove client-private `_meta` fields from MCP results returned to Code Mode.
- Include the new setting in the configuration schema and MCP config
serialization.
## Testing
- Cover every exposure combination across Code Mode, Code-Mode-only sessions,
tool search, direct-only namespaces, and prefixed and unprefixed MCP names.
- Verify direct and nested execution paths, parallel-call support, config
round-tripping, and `_meta` filtering.
GitOrigin-RevId: 12dfcb78bb5c5ecf4d70f38a8b5022792463a27f
## Why
Concurrent tests without a tracing subscriber can cache the network policy
callsite as disabled, causing event-capture tests to miss events.
## What changed
- Attach the test event collector directly to the captured future.
- Rebuild the tracing interest cache before running that future while keeping
tracing out of its single-subscriber fast path.
GitOrigin-RevId: 20d6160d26f56ad9fd86afd5107cb4313a5fce73
## What changed
- Replace the ordering enum and `include_disabled` boolean with named
`ConfigLayerStack` iterators for each precedence direction.
- Keep enabled-only iteration separate from `all_layers_*` iteration, which
includes disabled layers.
- Update config consumers to use the iterator matching their precedence and
disabled-layer requirements without allocating an intermediate `Vec`.
## Testing
- Add coverage that verifies ordering and disabled-layer filtering for all four
iterators.
GitOrigin-RevId: d14df3db4a2eae80ba97cdec246bff405e6c5f3d
## What changed
- Find the highest-precedence user layer by scanning the ordered config layers in reverse.
- Remove the cached user-layer index and its bookkeeping when constructing or updating a layer stack.
- Keep layer-order validation focused on validating ordering.
GitOrigin-RevId: e20510b868224ce25b2d2179f07d262d6b36e976
## Why
Codex Apps tool catalogs can exceed the standard MCP limit of 2,048 items and
still need to remain available through tool search.
## What changed
- Allow up to 8,192 catalog items for host-owned `codex_apps` registrations.
- Keep the 2,048-item limit for other MCP servers, including extensions that
register the `codex_apps` name.
- Restrict the shared Codex Apps tool cache to host-owned registrations.
## Testing
- Cover large catalogs for built-in and hosted-plugin registrations while
verifying that another extension retains the standard limit.
- Verify that a same-named extension cannot read the host-owned Apps cache.
GitOrigin-RevId: 2265693a460e9be63c4c849169cba7324def4f49
## Why
Managed proxy helpers can keep captured standard streams open after the sandboxed command exits. Proxy readiness can also fail when inherited standard descriptors are already closed, and zombie owners can leave stale socket directories behind.
## What changed
- Detach bridge and cleanup-worker standard I/O to `/dev/null`.
- Move readiness pipe descriptors above the standard descriptor range.
- Treat zombie processes as exited when cleaning proxy socket directories.
- Move helper lifecycle handling into a dedicated module.
## Testing
Add coverage for output release after command exit, readiness with closed standard descriptors, zombie detection, and stale socket cleanup.
GitOrigin-RevId: 9f4081cba1b73442f02a895473c0383a795fa30e
## What changed
- Restart the test app server before injecting the source thread's active goal so a late idle hook cannot continue it.
- Remove the unrelated ordinary-fork assertion and keep the scenario focused on deferred goal continuation across paginated forks.
GitOrigin-RevId: 5383f56ffa1687a4ebd18b5e37cbd61290fa59a7
## What changed
- Enable `test_sync_tool` for the startup prewarm test.
- Hold the first user-turn response open while Git enrichment completes, then
verify the refreshed workspace metadata on the follow-up request.
GitOrigin-RevId: dd4254fba1ac2f346022988cc21a78648a29556a
## What changed
Derive code-mode tool names directly from registered tool names when checking
eligibility and collisions. Skip unsupported and empty tool specs explicitly,
without serializing schemas or augmenting descriptions that are not used during
executor registration.
GitOrigin-RevId: b439327aafd85ce3745d900301b7e48b56dbd6b4
## What changed
- Allow up to five seconds for focus-triggered input to appear in the PTY test.
- Check for unexpected foreground and background palette queries while waiting for both immediate and delayed input, so either focus path fails with the palette-specific error.
GitOrigin-RevId: b8ed4f1ced257865140b1518c8959f65371e744f
## What changed
Move shell snapshot validation in the intercepted `apply_patch` test until after
the turn completes and its patch lifecycle events have been collected.
GitOrigin-RevId: 1c16bcfcc97f09165c4a2eb2d15f4059a14b2a99
## What changed
- Route verified patches from both direct tool calls and intercepted shell commands through a shared execution helper.
- Simplify patch safety preparation to return either a runtime invocation or a rejection directly.
- Keep permission resolution, approval handling, event emission, runtime execution, and diff tracking in the common path.
GitOrigin-RevId: e16aae8e91d96b8108aefac7f3fbf2da2cbac016
## What changed
Make `ExtensionRegistryBuilder` wrap an `ExtensionRegistry` and register
contributors directly on it. Return the wrapped registry from `build()` instead
of copying each field into a new value.
GitOrigin-RevId: ae04de99c17ea951156f9892151b4196c70026bf
## What changed
Replace the type-erased `FragmentRegistration` trait and proxy objects with a
static list of `matches_text` function pointers. Remove the registration types
and their re-exports while preserving the existing contextual fragment matchers.
GitOrigin-RevId: 4e5a296b278984c6e783a67f48fa4bb37646946b
## What changed
- Store each turn's host skills snapshot in `ExtensionData` and expose it through
`TurnContext::skills_snapshot`.
- Lazily keep implicit skill invocation deduplication state in the same turn-scoped
store.
- Carry the skills snapshot into review turns without a separate
`TurnSkillsContext` field.
GitOrigin-RevId: d71579b5d230bd0b5f3c2ad062861f74d614d88c
## What changed
- Share argument normalization, server selection, cursor validation, and pagination setup between `list_mcp_resources` and `list_mcp_resource_templates`.
- Use a generic server-tagged wrapper to build single-server and deterministically sorted all-server payloads for both resources and templates.
## Testing
- Cover shared argument normalization and sorted resource-template payload serialization.
GitOrigin-RevId: 3505e4c1d2b95d8bbbb01f6f930fb914309eb898
## What changed
- Increase the listener, stream read, and overall test timeouts from 2 to 10 seconds.
- Switch the accepted stream back to blocking mode before reading the request.
GitOrigin-RevId: 1e876350421d7791194e145751603c4d709c7d39
## What changed
- Emit dynamic tool-call analytics for code mode `exec` and `wait` calls, including duration and terminal status.
- Associate code-mode cells and their nested tool calls with the originating and subsequent model response IDs.
- Buffer correlated tool events until the next response is known, while flushing pending events when turns, threads, or the analytics queue close.
## Testing
- Add reducer coverage for correlating `exec`, `wait`, and nested tool events across turns.
- Add an app-server test that verifies production analytics include the cell and response IDs for a code-mode `exec` call.
GitOrigin-RevId: 8b34ab59c1ad68ebbd57f9c8a05f19ab04df48a4
## What changed
- Store synchronization primitives directly in `TurnMetadataState` instead of
wrapping each field in its own `Arc`.
- Require an `Arc<TurnMetadataState>` when spawning Git enrichment so the
background task retains the complete state.
- Update Git enrichment tests to use the shared state container.
GitOrigin-RevId: 745ca6504c0a6533cb7726d045905472e51726df
## What changed
- Add a shared runner for MCP resource operation lifecycle events, output serialization, truncation, timing, and error handling.
- Use it for listing resources, listing resource templates, and reading resources.
GitOrigin-RevId: 84cae2e01a096d5b8ed97ea1cb462f01fe2ed1f9
## What changed
- Charge `codex_rollout_budget_units` against the shared rollout budget when the provider includes it in response usage.
- Fall back to weighted input and output token accounting when provider units are absent.
- Reject non-finite or negative provider units as a fatal response error.
## Testing
- Cover provider units in reminder thresholds and local and remote compaction budget exhaustion.
- Verify invalid units fail without retrying the response.
GitOrigin-RevId: b452403e365985854d16f298d1ba46383e9892c4
## What changed
- Parse `codex_rollout_budget_units` from completed Responses API usage into
`TokenUsage`.
- Keep the provider-only value out of serialized protocol, JSON schema, and
TypeScript representations.
## Testing
- Cover the value in unit and end-to-end SSE response parsing tests.
GitOrigin-RevId: a53d4202beb9c8985d25894cb10e7c01f20b2a44
## What changed
- Accept the allowlisted `.onboarding_entrypoint=life_sciences` suffix on a valid OAuth state while continuing to reject unknown or malformed suffixes.
- Return the parsed callback metadata from the login server without including the suffix in the token exchange redirect URI.
- Add the optional `onboardingEntrypoint` field and `life_sciences` value to the v2 `account/login/completed` notification and generated protocol exports.
## Testing
- Cover valid, invalid, and repeated OAuth state suffixes.
- Verify the end-to-end login callback result, token redirect URI, and app-server notification payload.
GitOrigin-RevId: c2ee146328ddd43c03a503260d5d492e186fa889
## Why
Setting or clearing a thread goal could reconcile an already indexed rollout and overwrite SQLite-only thread metadata, including the thread preview.
## What changed
Skip rollout reconciliation when SQLite already references the same existing rollout and its session metadata matches the requested thread. Continue reconciling when the SQLite row is missing or the stored rollout is invalid so goal mutations can still repair thread metadata.
## Testing
Add coverage that verifies goal set and clear preserve SQLite previews, and that goal set restores a deleted SQLite thread row.
GitOrigin-RevId: cf19cef98559fc67182ac7430e4e50ed2eed83f3
## Why
Agent Plugins use a schema-declared root `plugin.json` and can have dotted names or versions that do not fit Codex's directory-safe version format. The packaging and installation paths still assumed the legacy manifest layout and identifier rules.
## What changed
- Recognize valid root Agent Plugin manifests when discovering, packing, and installing plugins, while leaving unrelated root manifests on the legacy path.
- Accept safe dotted plugin names, default missing Agent Plugin versions to `1.0.0`, and derive stable directory-safe versions when necessary without rewriting the portable manifest.
- Skip legacy command migration for Agent Plugins and reject symlinks or other unsupported file types while copying plugin sources.
## Testing
Add coverage for portable bundle round trips, manifest discovery, dotted names, version handling, command preservation, and symlink rejection.
GitOrigin-RevId: 61476c4c4100495842253d8b429c0b896490962d
## What changed
Increase the maximum number of items collected across paginated MCP tool,
resource, and resource-template discovery requests from 1,024 to 2,048.
GitOrigin-RevId: b6ce0691d6a217e92faaf141cbf09978b296eeda
## What changed
- Accept two-stroke bindings such as `ctrl-x ctrl-s` in TUI keymap configuration while preserving arrays as alternative bindings.
- Route chords through the active TUI context, show pending and configured chord hints, and cancel pending chords on `esc`, context changes, or a one-second timeout.
- Extend `/keymap` to capture, add, replace, and display key chords.
- Reject chords that conflict with active single-key bindings, overlapping-context chords, or reserved terminal shortcuts.
## Testing
- Add coverage for chord parsing, validation, matching, timeout and cancellation behavior, context dispatch, key capture, and rendered shortcut hints.
GitOrigin-RevId: dfdede03f518d4936ad8621ebf5537026f2b282e
## What changed
- Reattach recorded `executed_tool_calls` metadata when an output is included in a subsequent prompt.
- Bound retained metadata to 32 KiB, prioritizing recent calls and reporting omitted calls in truncation metadata.
- Drop retained entries after their corresponding outputs leave the prompt history.
## Testing
- Cover metadata replay, cleanup after compaction, bounded retained history, and propagation through later tool requests.
GitOrigin-RevId: 1c23a26123be3b7ad51c61f4ad522139b71bb773
## What changed
- Raise the maximum remote plugin bundle download from 50 MiB to 100 MiB.
- Raise the maximum total extracted bundle size from 250 MiB to 512 MiB.
GitOrigin-RevId: 17ff4c256dd496afa7fd08d1e8e4cdbd40d49500
## What changed
- Carry dimensions on resize events and reuse the cached screen size for ordinary draws.
- Refresh terminal geometry after resize settling, process resume, and external program execution.
- Pass the resolved size through transcript reflow, viewport updates, and standalone TUI screens so each frame uses consistent dimensions.
## Testing
- Cover cached draws, resize-driven rendering, settled-size rechecks, resume refreshes, and alternate-screen size updates.
GitOrigin-RevId: 9d83dbab50d7a038bd59b4c66c13abcac69ddc79
## What changed
- Move JSON-RPC request, notification, response, error, and malformed-message handling into a dedicated `RequestDispatcher`.
- Keep the connection loop responsible for receiving events and closing the connection when dispatch reports a terminal condition.
## Testing
- Add an integration test confirming that ordinary requests are processed serially by default, including when a blocking `process/read` queues later requests.
GitOrigin-RevId: 29d1358d4524edd492ff3855b29f23c42c8b3390
## What changed
- Add the optional `delegationAckFiller` field to `thread/realtime/start`.
- Forward explicit `true` or `false` values to V3 Frameless Bidi session payloads as `delegation.ack_filler`; leave the field absent when unspecified so the Realtime API default is preserved.
- Document that V1 and V2 ignore this setting.
## Testing
- Cover both boolean values in Frameless Bidi serialization and verify the WebRTC V3 session creation payload.
GitOrigin-RevId: 9fe8efd14e42738c3e86e3e9a8af81d0f4521244
## What changed
- Run `git init` to mark command execution in the affected pre-tool hook tests.
- Keep marker repositories in per-test temporary directories and assert against
their `.git` directories.
- Check `git version` output when a code-mode hook rewrites a nested command.
GitOrigin-RevId: 71c54ef525b08e1d592f69c8b88eed1ee6403e4f
## Why
Clients need to know whether a `request_user_input` request must wait for an
explicit response or may auto-resolve. Using `autoResolutionMs` as that signal
conflated the blocking decision with timeout policy.
## What changed
- Add required `isBlocking` fields to user input protocol and app-server
request payloads. Plan-mode requests are blocking, while requests from other
enabled modes are non-blocking.
- Drive the TUI's auto-resolution behavior from `isBlocking` and remove
`autoResolutionMs` from the model-facing tool schema.
- Deprecate `autoResolutionMs` while retaining it for compatibility, and treat
legacy payloads without `isBlocking` as blocking.
## Testing
- Cover mode-derived blocking behavior, legacy deserialization, app-server
forwarding, delegated requests, and TUI auto-resolution.
GitOrigin-RevId: 29aade657ef743065ec264376ba567a9b353d7d7
## What changed
- Implement `plugin/search` by querying the remote plugin service without using the catalog cache.
- Support global, workspace, and personal scopes with bounded page sizes and passthrough cursors.
- Respect plugin feature gates and omit shared workspace results when plugin sharing is disabled.
- Keep search terms and pagination tokens out of transport errors and telemetry, and return search results as uninstalled plugin summaries.
## Testing
- Add remote search coverage for request parameters, result conversion, authentication, pagination, scope mapping, and error redaction.
- Add app-server coverage for remote-plugin and plugin-sharing feature gates.
GitOrigin-RevId: ac29c5480ed8089d998b6275bc5f11c8d9a43fd1
## What changed
- Add optional `realtimeStartInstructions` and `realtimeEndInstructions` fields to `thread/realtime/start`.
- Apply the instructions when Codex enters and leaves realtime mode, while preserving the existing defaults when either field is omitted.
- Limit each field to 8,192 estimated tokens.
## Testing
- Cover custom entry and exit instructions across realtime state transitions.
- Cover token-limit validation for both fields.
GitOrigin-RevId: b9cfa1ff22e084d8a6b2ed56bd27d95ffdd6b6bd
## What changed
- Add the experimental `plugin/search` request with search term, scope, working-directory, cursor, and limit parameters.
- Define paginated results that include each plugin's marketplace name and optional local marketplace path.
- Route the request through app server while returning method-not-found until the search implementation is available.
## Testing
- Add an app-server test that verifies `plugin/search` returns JSON-RPC error `-32601` with the expected not-implemented message.
GitOrigin-RevId: 755400468331f0887f7da7d933195c347aadf6d0
## What changed
- Load `environments.toml` in one operation, falling back to the default
environment provider only when the read reports that the file is missing.
- Attempt the default daemon socket connection directly instead of checking the
socket path first.
- Preserve each file search entry's type from the directory walker, avoiding a
second metadata lookup when producing matches and correctly classifying
followed directory symlinks.
- Write atomic-file contents through the existing temporary-file handle rather
than reopening its path.
## Testing
- Add Unix coverage for classifying a followed directory symlink as a
directory in file search results.
GitOrigin-RevId: aa6739cac23953ca8b9a3fcb53e43e72c41cfdf9
## Why
Legacy thread histories did not use the cross-process writer ownership guard already applied to paginated histories.
## What changed
- Acquire and retain a writer lock whenever a legacy or paginated thread is created or resumed.
- Apply the same ownership checks when archiving or deleting either history mode.
- Release ownership when initialization fails or the active thread shuts down, allowing another process to resume it.
## Testing
- Cover competing create and resume operations for both history modes.
- Cover archive and deletion conflicts for legacy and paginated descendants.
GitOrigin-RevId: 1be79b01966ad2634ed79eb96c0aa0db8eafe2fa
## What changed
- Record the effective detail setting and source and prepared dimensions for
successfully decoded images.
- Associate message images with their role and tool-output images with their
originating call ID.
- Include the collected image preparation metadata in the corresponding turn
analytics event.
## Testing
- Cover image resize metadata, tool-output attribution, turn aggregation, event
serialization, and app-server analytics output.
GitOrigin-RevId: dbfb5ba1ce7cb02dd93d9b579c5109817f16aef5
