Haleclipse/codex rust-v0.154.0-cometix

0.154.0-cometix

on GitHub

New Features

• GPT-6-Astra is now available in the model picker and Amazon Bedrock catalogs. (openai#42879, openai#42619)
• Experimental worktree support lets you create isolated checkouts for new or forked sessions using --worktree or /worktree, then browse and resume them. (openai#42652, openai#43069, openai#43120, openai#43286)
• Answer questions inline while Codex continues working, using suggested choices or custom text without losing your main draft. (openai#42891, openai#42894, openai#42897)
• Windows sessions can now share a background Codex server, with daemon lifecycle commands and managed updates. (openai#42405, openai#42392)
• Vim editing gains R replace mode with undo and dot-repeat, plus more reliable Escape handling in legacy terminals. (openai#42194, openai#42584)
• Copying responses preserves formatting in rich-text apps, and /copy can copy status output or individual session fields. (openai#42847, openai#43055)

Bug Fixes

• Existing sessions pick up newly installed plugin tools and refresh skills and hooks after external plugin upgrades or rollbacks. (openai#42284, openai#42593, openai#42990)
• MCP connections coordinate OAuth token refreshes and surface login challenges when refresh fails, without automatically replaying rejected tool calls. (openai#42413, openai#42552)
• Startup avoids running workspace-controlled helpers before trust is established, and the macOS sandbox blocks terminal input injection. (openai#42324, openai#42590)
• Remote resume and fork operations preserve saved permissions; fresh sessions and forks respect server model defaults unless explicitly overridden. (openai#43330, openai#43177, openai#43355)
• Resuming a conversation open in another app now shows a read-only transcript with a retry option while preserving your draft. (openai#43253)
• Automatic approval reviews better preserve authorization context through compaction and reject approvals invalidated by new user instructions or answers. (openai#42844, openai#42852, openai#43442)

Documentation

• Updated the bundled OpenAI Docs skill with GPT-6-Astra migration, compatibility, and prompting guidance. (openai#42931)

Chores

• The deprecated codex mcp-server entry point is no longer available. (openai#42993)

Changelog

Full Changelog: openai/codex@rust-v0.153.0...rust-v0.154.0

• openai#42188 Fix punctuation in npm packaging documentation @copyberry
• openai#42192 Use native spawning for bare macOS MCP commands @copyberry
• openai#42194 Add Vim replace mode to the TUI composer @copyberry
• openai#42196 Add managed worktree creation @copyberry
• openai#42199 Refactor shared TUI input routing @copyberry
• openai#42202 Separate TUI preferences from server configuration @copyberry
• openai#42204 Add macOS voice runtime projection @copyberry
• openai#42207 Retry TUI reconnects while threads are closing @copyberry
• openai#42208 Add GNU Linux voice runtime preparation @copyberry
• openai#42209 Add Windows voice runtime preparation @copyberry
• openai#42247 Track history notes thread hint outcomes @copyberry
• openai#42256 Skip Guardian scoring in User approval mode @copyberry
• openai#42270 Report the exec-server release version in environment info @copyberry
• openai#42284 Refresh plugin skills after out-of-process version changes @copyberry
• openai#42288 Fetch rules_rs zlib packages from Ubuntu snapshots @copyberry
• openai#42290 Expand Guardian history coverage across resume and rollback @copyberry
• openai#42293 Preserve verified answers across history compaction @copyberry
• openai#42298 Preserve retained answers across steer rollbacks @copyberry
• openai#42306 Stabilize the detached exec-server session resume test @copyberry
• openai#42309 Separate Windows sandbox provisioning from ACL refresh @copyberry
• openai#42314 Preserve target-native cwd in permission approval requests @copyberry
• openai#42316 Refactor exec-server startup futures @copyberry
• openai#42318 Support packaged managed Codex binary paths @copyberry
• openai#42319 Show live context compaction status in the TUI @copyberry
• openai#42320 Make the app-server thread unload delay configurable @copyberry
• openai#42324 Avoid executing PATH helpers before workspace trust @copyberry
• openai#42325 Render completed assistant messages directly during replay @copyberry
• openai#42326 Harden Windows control socket rendezvous @copyberry
• openai#42328 Support durable reasoning configuration updates @copyberry
• openai#42330 Protect Windows sandbox binaries from inherited write access @copyberry
• openai#42332 Package prepared runtimes with the voice host @copyberry
• openai#42334 Add a Windows sandbox provisioning protocol @copyberry
• openai#42337 Add an authenticated Windows sandbox provisioning client @copyberry
• openai#42341 Add Windows sandbox service lifecycle scaffolding @copyberry
• openai#42342 Harden Windows sandbox provisioning file handling @copyberry
• openai#42344 Prepare managed policy validation for Windows sandbox provisioning @copyberry
• openai#42348 Add Windows sandbox client authentication @copyberry
• openai#42351 Enable authenticated Windows sandbox provisioning @copyberry
• openai#42353 Add experimental Windows sandbox service provisioning @copyberry
• openai#42354 Add free-form asynchronous user messages @copyberry
• openai#42356 Initialize questions in buffered replay test messages @copyberry
• openai#42358 Extend rate limit reads with usage capabilities @copyberry
• openai#42364 Support graceful daemon shutdown on Windows @copyberry
• openai#42366 List managed worktrees for a repository @copyberry
• openai#42369 Keep SQLite history projection moving past invalid records @copyberry
• openai#42370 Improve MCP server startup error logging @copyberry
• openai#42372 Add Luna Reserve usage fallback to the TUI @copyberry
• openai#42373 Add attributed exec process lifecycle telemetry @copyberry
• openai#42374 Extract PID startup into a dedicated module @copyberry
• openai#42375 Clean up Windows sandbox resources on app uninstall @copyberry
• openai#42377 Make app-server realtime sessions always available @copyberry
• openai#42378 Route rollout reads through the canonical JSON decoder @copyberry
• openai#42380 Require confirmation for safety-buffered retries @copyberry
• openai#42381 Support managed app-server lifecycle on Windows @copyberry
• openai#42383 Update rmcp to 3.2.0 @copyberry
• openai#42384 Add an RMCP OAuth credential store adapter @copyberry
• openai#42385 Add experimental context management activation @copyberry
• openai#42386 Expose loaded thread environments in app-server responses @copyberry
• openai#42388 Recover deferred environments after provisioning failure @copyberry
• openai#42391 Authorize apply_patch in the executor path context @copyberry
• openai#42392 Support managed daemon updates on Windows @copyberry
• openai#42395 Expose the Codex version to commands and turn metadata @copyberry
• openai#42397 Extract focused TUI logic into submodules @copyberry
• openai#42399 Preserve restored input after resolved misalignment errors @copyberry
• openai#42401 Discover TUI collaboration modes from the app server @copyberry
• openai#42403 Expose the last accepted environment ready report @copyberry
• openai#42404 Read voice helper frames independently of pipe chunks @copyberry
• openai#42405 Support the app-server daemon on Windows @copyberry
• openai#42406 Honor explicit plugin mentions during MCP startup @copyberry
• openai#42408 Harden embedded composer input handling @copyberry
• openai#42410 Allow reviewing and continuing misalignment-paused chats @copyberry
• openai#42413 Enable coordinated MCP OAuth refresh @copyberry
• openai#42417 Expose managed application network requirements @copyberry
• openai#42419 Add session resume to the agent command center @copyberry
• openai#42422 Honor model requirements in Guardian computer-use scoring @copyberry
• openai#42425 Discover TUI experimental features from the server @copyberry
• openai#42428 Use the shared composer in the agent command center @copyberry
• openai#42432 Box the TUI resume picker future @copyberry
• openai#42445 Include originator in plugin measurement analytics @copyberry
• openai#42451 Acknowledge pending TUI steers by submission ID @copyberry
• openai#42453 Discover permission profiles from the app server @copyberry
• openai#42455 Show live task details in the agent command center @copyberry
• openai#42458 Expose thread originators through the app-server API @copyberry
• openai#42529 Register the Guardian thread context feature flag @copyberry
• openai#42552 Preserve MCP authentication challenges on tool calls @copyberry
• openai#42577 Preserve target-native paths in command approvals @copyberry
• openai#42579 Persist verified user answers in Guardian thread context @copyberry
• openai#42584 Recover Vim escape input in legacy terminals @copyberry
• openai#42588 Require Guardian review for incompatible compaction checkpoints @copyberry
• openai#42590 Harden the macOS sandbox against terminal input injection @copyberry
• openai#42593 Reload user config after local plugin installation @copyberry
• openai#42596 Record Windows sandbox private desktop usage @copyberry
• openai#42598 Report MCP tool discovery errors in server status @copyberry
• openai#42602 Deprecate detached review delivery @copyberry
• openai#42603 Expose global metrics installation in codex-otel @copyberry
• openai#42606 Support trusted headers for remote exec WebSockets @copyberry
• openai#42607 Add GPT-6-Astra to the bundled model catalog @copyberry
• openai#42609 Condense TUI startup warnings @copyberry
• openai#42619 Add GPT-6-Astra to Amazon Bedrock catalogs @copyberry
• openai#42623 Bound Noise handshakes by the exec server initialization timeout @copyberry
• openai#42624 Centralize prompt image detail modes @copyberry
• openai#42631 Initialize the packaged GStreamer runtime in the voice host @copyberry
• openai#42634 Add an injectable attachment store to ThreadManager @copyberry
• openai#42638 Update GPT-6-Astra Fast tier speed description @copyberry
• openai#42639 Warn when saved model defaults are overridden @copyberry
• openai#42640 Harden TUI parsing of assistant markup @copyberry
• openai#42641 Restore the inline TUI after full-screen overlays @copyberry
• openai#42650 Render assistant file citations as local links @copyberry
• openai#42652 Add managed worktrees to codex exec @copyberry
• openai#42654 Update the stable exec-server test to Codex 0.153.1 @copyberry
• openai#42657 Use a generic fallback model name in status tests @copyberry
• openai#42667 Tailor TUI cyber refusal notices to Daybreak eligibility @copyberry
• openai#42668 Cancel remote control enrollment on stdio shutdown @copyberry
• openai#42671 Preserve TUI sessions while starting replacement threads @copyberry
• openai#42674 Persist server-advertised experimental features from the TUI @copyberry
• openai#42676 Add WebRTC negotiation to the voice host @copyberry
• openai#42677 Narrow async user message guidance @copyberry
• openai#42682 Fix the worktrees experimental feature test fixture @copyberry
• openai#42716 Allow trusted symlinks beneath CODEX_HOME on macOS @copyberry
• openai#42718 Gate unified exec TTY support behind a feature flag @copyberry
• openai#42741 Make the TUI symlink startup test Bazel-compatible @copyberry
• openai#42744 Honor model-provided Guardian review policies @copyberry
• openai#42746 Handle pending network reviews after process completion @copyberry
• openai#42749 Improve automatic thread naming in the TUI @copyberry
• openai#42752 Preserve response IDs for fast collaborator tool events @copyberry
• openai#42755 Stabilize the interactive tmux startup safety test @copyberry
• openai#42758 Propagate response tickets to Guardian reviews @copyberry
• openai#42762 Retain user instructions in guardian thread context @copyberry
• openai#42767 Avoid port races in streamable HTTP tests @copyberry
• openai#42770 Preserve acceptance order in retained thread context @copyberry
• openai#42773 Avoid holding metadata permit during cold resume config load @copyberry
• openai#42781 Add direct SigV4 transport to exec-server @copyberry
• openai#42791 Keep TUI prompt history tied to local settings @copyberry
• openai#42792 Extract the note input view into its own module @copyberry
• openai#42798 Add data-use disclosures to the user report dialog @copyberry
• openai#42801 Keep the Windows sandbox command runner hidden @copyberry
• openai#42807 Add request-scoped Guardian approval decisions @copyberry
• openai#42811 Enable staging login issuer overrides in packaged builds @copyberry
• openai#42814 Support custom report event titles @copyberry
• openai#42819 Route Guardian approvals independently of async scoring @copyberry
• openai#42821 Report managed filesystem policy in codex doctor @copyberry
• openai#42823 Expose managed WebMCP policy through the app server @copyberry
• openai#42824 Refine user input guidance for GPT-6 @copyberry
• openai#42832 Preserve root authorization context in Guardian reviews @copyberry
• openai#42833 Preserve SystemRoot for Windows sandbox wrapper setup @copyberry
• openai#42835 Preserve Windows managed deny reads in the sandbox CLI @copyberry
• openai#42836 Make GPT-6-Astra user input guidance conditional @copyberry
• openai#42838 Preserve executor paths in Guardian approval reviews @copyberry
• openai#42841 Add a native Windows MXC sandbox adapter @copyberry
• openai#42842 Add Astra sparkle effects to the TUI composer @copyberry
• openai#42844 Retain user instructions in Guardian context @copyberry
• openai#42847 Preserve Markdown formatting when copying TUI responses @copyberry
• openai#42850 Use jemalloc for Linux musl binaries @copyberry
• openai#42852 Harden Guardian reviews after context compaction @copyberry
• openai#42854 Persist Daybreak preferences in thread metadata @copyberry
• openai#42863 Preserve precedence across feature requirement aliases @copyberry
• openai#42870 Avoid redundant filesystem sandbox path resolution @copyberry
• openai#42879 List GPT-6-Astra in the model picker @copyberry
• openai#42883 Add client-side exec-server RPC attempt metrics @copyberry
• openai#42889 Add TUI building blocks for inline async question editing @copyberry
• openai#42891 Integrate asynchronous questions into the TUI @copyberry
• openai#42894 Support selectable answers for asynchronous TUI questions @copyberry
• openai#42897 Add inline Other answers to async question choices @copyberry
• openai#42900 Establish root turn identity for independent tasks and memory requests @copyberry
• openai#42903 Preserve TUI question state and integrate history and queue navigation @copyberry
• openai#42904 Use static instructions for the Default collaboration mode @copyberry
• openai#42931 Update OpenAI Docs skill guidance for GPT-6 Astra @copyberry
• openai#42933 Wait for turn analytics before shutting down the Guardian v2 test @copyberry
• openai#42990 Refresh session hooks after external plugin updates @copyberry
• openai#42993 Remove the deprecated codex mcp-server command @copyberry
• openai#43000 Preserve the resolved multi-agent version when reverting threads @copyberry
• openai#43002 Replace Guardian tickets with parent response IDs @copyberry
• openai#43005 Add Guardian V2 failure reasons and connection timing metrics @copyberry
• openai#43031 Keep refreshed MCP tool catalogs with their clients @copyberry
• openai#43039 Refresh live thread tools through app/installed @copyberry
• openai#43043 Avoid filesystem scans when seeding the agents overview @copyberry
• openai#43055 Allow /copy to copy status output and individual fields @copyberry
• openai#43069 Support managed worktrees for interactive sessions and forks @copyberry
• openai#43070 Clarify comments in CI setup and the Rust workflow @copyberry
• openai#43074 Show a retryable error when the apps popup fails to load @copyberry
• openai#43079 Add opt-in local audio devices to the voice helper @copyberry
• openai#43083 Supply Bazel-managed CMake and Ninja for the bundled Opus build @copyberry
• openai#43090 Send processed microphone audio over RTP in voice-host @copyberry
• openai#43097 Add a helper-backed realtime WebRTC session API @copyberry
• openai#43099 Add receipt-verified native voice SDK export @copyberry
• openai#43100 Add bounded incoming Opus RTP handling to the voice host @copyberry
• openai#43102 Include GIO in voice SDKs and native runtimes @copyberry
• openai#43104 Move Guardian thread context into guardianv2 configuration @copyberry
• openai#43109 Add explicit toolchain inputs for native voice builds @copyberry
• openai#43110 Record reasoning effort changes in conversation history behind a flag @copyberry
• openai#43111 Add a Bazel target for native voice dependencies @copyberry
• openai#43113 Save subagent and memory opt-ins through the app server @copyberry
• openai#43114 Add Bazel preparation for native voice runtimes @copyberry
• openai#43117 Link Unix Bazel bindings against the prepared voice runtime @copyberry
• openai#43120 Add managed worktree creation to TUI session commands @copyberry
• openai#43121 Require a prepared runtime when assembling voice helper packages @copyberry
• openai#43125 Add explicit Windows tool selection for native voice builds @copyberry
• openai#43126 Expose native Windows build tools through Bazel targets @copyberry
• openai#43144 Add Windows MSVC Bazel targets for native voice libraries @copyberry
• openai#43147 Gate experimental context by model capability at session startup @copyberry
• openai#43177 Use server model defaults for fresh TUI startup @copyberry
• openai#43178 Allow guarded legacy resume with background migration enabled @copyberry
• openai#43244 Add bounded GStreamer playback components to the voice host @copyberry
• openai#43248 Connect voice-host RTP audio to speaker playback @copyberry
• openai#43253 Show read-only conversations when resume encounters an active writer @copyberry
• openai#43261 Use server defaults when starting TUI background tasks @copyberry
• openai#43265 Add experimental user verification API contracts @copyberry
• openai#43279 Include linked worktrees in TUI session discovery @copyberry
• openai#43281 Move npm package staging into a separate release workflow job @copyberry
• openai#43282 Make Bazel binary stamping opt-in @copyberry
• openai#43286 Add a managed worktree browser to the TUI @copyberry
• openai#43289 Add capability-gated MCP user-verification handling @copyberry
• openai#43298 Defer managed worktree transitions to fresh TUI loop iterations @copyberry
• openai#43304 Isolate Bazel build commit metadata from Rust compilation inputs @copyberry
• openai#43308 Replace Windows app-server shutdown files with socket requests @copyberry
• openai#43315 Resolve session labels uniquely before acting on them @copyberry
• openai#43325 Sort JSON schema object keys for consistent Cargo and Bazel output @copyberry
• openai#43330 Preserve saved permissions when resuming or forking remote tasks @copyberry
• openai#43340 Enable remote named permission profile selection in the TUI @copyberry
• openai#43352 Add opt-in MCP user-verification transport @copyberry
• openai#43355 Let the app server resolve implicit model settings for CLI forks @copyberry
• openai#43359 Show the server's model provider ID in TUI status @copyberry
• openai#43360 Use app-server metadata for TUI session restoration @copyberry
• openai#43376 Defer resume picker and directory changes to a fresh TUI stack @copyberry
• openai#43408 Avoid WebSocket connection waits in Guardian v2 classification @copyberry
• openai#43419 Initialize cwd in TUI resume and fork test fixtures @copyberry
• openai#43421 Remove the app-server README and its contributor guidance references @copyberry
• openai#43423 Remove the app-server docs update requirement from AGENTS.md @copyberry
• openai#43426 Handle Luna HTTP requests in guardian history tests @copyberry
• openai#43428 Notify opted-in stdio MCP servers of auth changes @copyberry
• openai#43432 Route approvals through the extension decision API @copyberry
• openai#43442 Keep Guardian review evidence consistent and reject stale approvals @copyberry
• openai#43444 Pin V8 release manifests and prevent published release replacement @copyberry
• openai#43447 Route MCP elicitations through the shared approval decision path @copyberry
• openai#43454 Add diagnostic labels to shell snapshot capture metrics @copyberry
• openai#43456 Wait for thread idle before rollback in model-switching tests @copyberry
• openai#43458 Centralize Guardian context mode and checkpoint policy @copyberry
• openai#43462 Remove legacy Guardian approval review paths @copyberry
• openai#43472 Recover missing Guardian root instructions in acceptance order @copyberry
• openai#43478 Retain inherited Guardian instructions in standalone forks @copyberry
• openai#43490 Expose shared Guardian reviewer helpers through guardian_review @copyberry
• openai#43491 Include unloaded children in multi-agent v2 environment context @copyberry
• openai#43494 Limit archive rollout reads to requested threads @copyberry
• openai#43495 Allow internal sessions to fork from selected history @copyberry
• openai#43504 Treat zombie processes as inactive in the Unix PID backend @copyberry
• openai#43513 Expose a stable executor build identity in environment metadata @copyberry
• openai#43519 Set recursion_limit to 256 for app-server, exec, and TUI @copyberry