Codex CLI v0.168.0 Release Notes — What's New

New Features

Browse model providers under Sessions → Model providers and choose the default for new chats. (#59541, #59722)
Enable features.tui_workspace to view a side conversation beside the main chat. Drag the divider to resize the panes or press leader z to zoom the focused pane. (#58992, #59527, #59595)
Attach recordings when using an audio-capable model. Codex skips confirmed rejected recordings on later turns and uses clip duration to keep more short recordings in context. (#59234, #59268)
Enable features.image_generation_animation to generate or edit animated GIFs with optional frame count and timing controls. (#59073)
Fullscreen Codex displays image previews and pets in tmux on Ghostty and Kitty with allow-passthrough all. (#59579, #59580)
Set tui.notification_method = "osc777" for terminal notifications. Approval notifications show longer command previews and distinguish terminal input from network access requests. (#59256)

Bug Fixes

Local TUI history and resume pickers filter newly created threads by the signed-in ChatGPT account. (#58918)
Open and resumed chats use their saved model provider after you change the default for new chats. (#59503)
The TUI renders replies interrupted by tool calls as a single Markdown message and restores reasoning summaries when replaying a conversation. (#59698, #59700)
Steering messages reach running turns when sent during the final transcript save or after switching back to a long-running chat. (#58877, #58998)
Fuzzy file search gives filename matches more weight so matching parent directories are less likely to crowd them out. (#59709)
Idle local MCP servers on Windows leave filesystem I/O available to the rest of Codex. (#59726)

Documentation

The app-server reference covers listing model providers, creating custom gateways, and updating gateway, Bedrock, and GCP settings. (#58845, #58876)

Chores

Codex warns when features.multi_agent is explicitly configured and points to agents.enabled for disabling subagent tools. (#59479)
Contributors can run Bazel CI jobs locally without BuildBuddy credentials. (#59055)

Changelog

Full Changelog: https://github.com/openai/codex/compare/rust-v0.167.0...rust-v0.168.0

#58766 Distinguish update handoffs in TUI startup telemetry @etraut-openai
#58767 Validate /fork before consuming the composer draft @etraut-openai
#58829 Accept agent follow-ups through native turn input delivery @won-openai
#58845 Add model provider discovery to the app server @alexsong-oai
#58876 Add model provider configuration writes to the app-server @alexsong-oai
#58877 Handle steers accepted during the final transcript flush @charliemarsh-oai
#58886 Preserve timestamp ties when paginating threads @jif-oai
#58890 Add a feature-gated account check for CLI thread resume @jif-oai
#58894 Support model catalog overrides for plugin guidance @goodman-oai
#58899 Add native patch support to direct Guardian review @swastik-oai
#58903 Support permission requests in direct Guardian review @swastik-oai
#58904 Add a standalone HTTP MCP server for the agent message board @jif-oai
#58908 Package the message board MCP connection as a plugin @jif-oai
#58909 Add an in-memory launcher for the message-board MCP server @jif-oai
#58918 Scope local history discovery to the current ChatGPT account @jif-oai
#58927 Fix live stream reflow at the allocated conversation width @fcoury-oai
#58931 Support direct Guardian review of patches invoked through exec_command @swastik-oai
#58933 Release message-board state before HTTP shutdown returns @jif-oai
#58935 Defer Guardian context trimming in elastic preparation @ioanbudea-OAI
#58936 Extract model-provider unit tests into separate files @ioanbudea-OAI
#58937 Verify queued network approval wakeups before polling again @seanh-oai
#58938 Skip attribution for direct MCP tool calls @victor-openai
#58939 Refine additional-context byte histogram buckets @akira-oai
#58940 Retire the daemon_auto_start feature flag @etraut-openai
#58945 Fix model selection after provider setup @etraut-openai
#58946 Preserve Guardian sender context for cloud deliveries with orbit IDs @felixxia-oai
#58950 Recover WebSocket transport after transient HTTP fallback @charliemarsh-oai
#58953 Support caller-owned sideband channels for realtime attachments @jiahao-oai
#58954 Add a thread-store builder with host cleanup support @felixxia-oai
#58955 Delete Guardian evidence boards when permanently deleting threads @felixxia-oai
#58956 Allow volatile World State sections to refresh after compaction @rhan-oai
#58957 Cache marketplace name mismatches during automatic upgrades @charliemarsh-oai
#58959 Recover inherited history from alternate rollout copies @charliemarsh-oai
#58960 Prefer indexed rollout paths when resolving ancestors @charliemarsh-oai
#58962 Preserve Guardian budget policy through review recovery @ioanbudea-OAI
#58963 Add host-owned turn admission to the agent runtime @rhan-oai
#58964 Stabilize and enable client tool output retention by default @jif-oai
#58966 Move protocol permissions tests into a separate file @viyatb-oai
#58969 Avoid unnecessary rendering during TUI history loading and insertion @etraut-openai
#58970 Keep automatic thread title saves off the TUI event loop @etraut-openai
#58971 Fix Escape handling in memory reset confirmation @imac-oai
#58973 Add opt-in MCP discovery for in-memory agent message boards @jif-oai
#58974 Bind final-message output before running sandboxed tools @viyatb-oai
#58975 Add usage accounting for host-owned agent runtimes @rhan-oai
#58978 Enable the segment heap on Windows CI runners @anp-oai
#58980 Record turn critical path metrics and trace fields @bkotsopoulos
#58983 Extend and bound exec-server session recovery @richardopenai
#58984 Allow hosts to restrict a thread's root-only capabilities @rhan-oai
#58986 Move agent message-board backend construction into the extension @jif-oai
#58988 Respect delivery boundaries for injected agent messages @won-openai
#58989 Keep HTTP fallback active for the rest of the session @charliemarsh-oai
#58990 Preserve modified Ctrl+C keys in active TUI views @imac-oai
#58992 Add opt-in side conversation panes to the TUI @fcoury-oai
#58996 Deliver environment hook decisions while output reads are queued @andrewgu-oai
#58998 Preserve steering when returning to a long-running turn @tamird
#59000 Allow hosted child threads to inherit a root session ID @rhan-oai
#59001 Test connection events against active environment selections @anp-oai
#59003 Preserve Guardian review framing under elastic context limits @ioanbudea-OAI
#59007 Add host cycling and saved defaults for new TUI conversations @bc-openai
#59009 Centralize filesystem permission resolution for sandbox enforcement @viyatb-oai
#59011 Add optional total token limits to elastic section configuration @ioanbudea-OAI
#59015 Make SQLite schema migrations safe for concurrent startup @dkovalenko-oai
#59016 Make observed thread metadata writes atomic and retryable @dkovalenko-oai
#59018 Add WebSocket connection metrics with reauthentication labels @bromano-oai
#59020 Preserve sandbox context in prepared MCP calls @sayan-oai
#59021 Allow concurrent readers of the installation ID @dkovalenko-oai
#59025 Remove strict third-party tool deferral in Code Mode Only @rishab-openai
#59028 Use prepared MCP owner permissions in Guardian reviews @sayan-oai
#59033 Add web search provider metadata and configuration @yitaoh-oai
#59034 Add model Search discovery to MCP bindings @yitaoh-oai
#59036 Resolve model-selected search providers with policy-aware fallback @yitaoh-oai
#59041 Expose assistant content token counts in completion notifications @aojiao-oai
#59043 Tie executor attachments to environment selection lifetimes @anp-oai
#59047 Decode live transcript updates in the cloud client @bc-openai
#59055 Run Bazel CI locally without BuildBuddy credentials @dkovalenko-oai

…