0.157.0-cometix

New Features

• Added GPT-6 Sol and Luna, including Amazon Bedrock support and migration prompts for older models. (openai#47332, openai#47347)
• Enabled fullscreen transcripts by default and added Shift-click to extend text selections. (openai#47178, openai#47414)
• Enabled automatic background-server startup for eligible interactive sessions, with recovery choices when server settings are incompatible. (openai#47179, openai#47318)
• Added an f shortcut to fork conversations open in another app, preserving drafts and queued prompts. (openai#47185)
• Made /import available in remote sessions and local background-server sessions. (openai#47317)
• Improved terminal rendering with Unicode bullets, checkboxes, aligned equations, and optimization notation. (openai#47191, openai#47322)

Bug Fixes

• Preserved active voice conversations when switching threads. (openai#47381)
• Recovered unsent question answers into the composer when turns end, without disrupting active history searches. (openai#47422, openai#47423)
• Respected tmux mouse settings and restored native scrollback for Terminal.app over SSH in automatic screen mode. (openai#47399, openai#47417)
• Fixed configured proxy routing for realtime connections and standalone web search, including search redirects. (openai#47101, openai#47142, openai#47204)
• Added retries for transient file-upload failures and increased the upload timeout to five minutes. (openai#47122, openai#47393)
• Enforced network restrictions across redirects and ongoing HTTP and WebSocket traffic, including cancellation when policy changes revoke access. (openai#47389, openai#47407)

Changelog

Full Changelog: openai/codex@rust-v0.156.0...rust-v0.157.0

• openai#47063 Add a partial index for agent message board root posts @jif-oai
• openai#47064 Stop redundant message-board reads when limits reach one @jif-oai
• openai#47065 Deduplicate deprecation notices in the TUI transcript @etraut-openai
• openai#47068 Highlight selected newlines in the transcript view @etraut-openai
• openai#47073 Wait for thread idle in Guardian context budget tests @jif-oai
• openai#47074 Refresh host-supplied cloud skills at turn startup @TAFOYA-OAI
• openai#47075 Include attributed previews in agent message board notifications @jif-oai
• openai#47077 Serialize environment updates and cancel removed pending attachments @sayan-oai
• openai#47079 Mask daemon socket paths exposed through ancestor bind mounts @nicksteele-oai
• openai#47081 Track cumulative MCP attribution across requests and thread history @peilin-openai
• openai#47082 Show the voice toggle in the TUI shortcut overlay @etraut-openai
• openai#47083 Extract per-thread enrichment into a dedicated async helper @aibrahim-oai
• openai#47084 Avoid building a host SQLite driver for SQLx macros @aibrahim-oai
• openai#47085 Update model catalog descriptions and GPT-5.6-Sol priority @andrewgu-oai
• openai#47086 Format analytics credit usage with consistent decimal precision @fcoury-oai
• openai#47088 Disable unused default dependency features @aibrahim-oai
• openai#47089 Centralize JSON-RPC response serialization through payloads @aibrahim-oai
• openai#47094 Restrict Unix local MCP servers to stdio descriptors @yuzhu-oai
• openai#47095 Remove the rust-release-prepare workflow @andrewgu-oai
• openai#47096 Keep TUI hints immediately above the composer @fcoury-oai
• openai#47100 Preserve assistant answers in bounded TUI task responses @etraut-openai
• openai#47101 Honor configured proxies for realtime WebSocket connections @celia-oai
• openai#47106 Preserve originating turn metadata for yielded skill calls @eddie-openai
• openai#47108 Preserve required Windows runtime variables for filesystem helpers @iceweasel-oai
• openai#47112 Lowercase the retry hint for conversations open in another app @fcoury-oai
• openai#47113 Persist thread creator identity in rollouts and SQLite @nornagon-openai
• openai#47114 Preserve and expose thread item lifecycle timestamps @chiam-oai
• openai#47116 Honor the configured product SKU in remote plugin requests @jessedu-oai
• openai#47118 Support model-catalog overrides for Code Mode tool messages @rhan-oai
• openai#47121 Pass thread IDs to attachment uploads @kchainani-oai
• openai#47122 Increase OpenAI file blob upload timeout from 60 seconds to 5 minutes @wilkes-oai
• openai#47125 Add extra policy configuration for Guardian reviews @won-openai
• openai#47129 Preserve foreign working directories in extension tool environments @anp-oai
• openai#47130 Remove the ultrafast service tier from gpt-5.6-sol @andrewgu-oai
• openai#47132 Support caller-provided MITM CAs in the network proxy @viyatb-oai
• openai#47137 Prevent horizontal transcript selection from triggering autoscroll @imac-oai
• openai#47142 Honor system proxy settings for standalone web search @celia-oai
• openai#47143 Extract exec-server CLI startup into a dedicated module @anp-oai
• openai#47149 Update code mode catalog snapshot hashes @andrewgu-oai
• openai#47155 Ignore stale thread-close notifications after resuming a thread @etraut-openai
• openai#47158 Harden gateway OAuth credential persistence and error redaction @alexsong-oai
• openai#47159 Expose HTTP origins in MCP server status @frantic-openai
• openai#47162 Apply workspace routing to Guardian v2 classifier requests @acrognale-oai
• openai#47170 Add explicit gateway login control and authentication status @alexsong-oai
• openai#47174 Preserve empty cell metadata under pending tool-call pressure @ningyi-oai
• openai#47178 Enable the fullscreen transcript by default @etraut-openai
• openai#47179 Enable automatic daemon startup by default @etraut-openai
• openai#47185 Allow forking conversations locked by another app in the TUI @fcoury-oai
• openai#47191 Render aligned equations and optimization notation in TUI math @etraut-openai
• openai#47199 Separate thread preparation from execution persistence @dermanyang-oai
• openai#47204 Resolve proxy routes for standalone web search redirects @celia-oai
• openai#47207 Add explicit gateway OAuth sign-in to app-server @alexsong-oai
• openai#47212 Preserve MCP request trace context across transport workers @bromano-oai
• openai#47235 Route multi-agent spawning through SpawnRequest @jif-oai
• openai#47236 Suppress agent message board notifications to the post author @jif-oai
• openai#47240 Prevent V2 agent eviction from racing with queued messages @jif-oai
• openai#47244 Centralize agent resume handling in LocalAgentControl @jif-oai
• openai#47248 Add explicit app and account targets to MCP resource reads @victor-openai
• openai#47249 Drop descendant channel posts from retained remote compaction v2 history @jif-oai
• openai#47252 Return agent snapshots from close_agent @jif-oai
• openai#47257 Preserve explicit message-board unsubscribes when posting @jif-oai
• openai#47258 Test spawn settings reporting after child removal @jif-oai
• openai#47259 Subscribe authors to message-board channels created by posting @jif-oai
• openai#47263 Expose hosted plugin extensions in app-server summaries @victor-openai
• openai#47267 Stream agent snapshots from local status subscriptions @jif-oai
• openai#47271 Implement AgentControl for LocalAgentControl @jif-oai
• openai#47272 Allow Guardian sync review across compaction hash differences @felixxia-oai
• openai#47275 Enable Guardian thread context by default @felixxia-oai
• openai#47281 Add regression coverage for message board unsubscribe persistence @jif-oai
• openai#47287 Update the Guardian extra-policy snapshot for retained instructions @felixxia-oai
• openai#47292 Consolidate agent operations in the AgentControl implementation @jif-oai
• openai#47301 Separate local agent runtime state from controller handles @jif-oai
• openai#47302 Use compact display metadata for Guardian review threads @charliemarsh-oai
• openai#47303 Clean up legacy Guardian thread metadata in SQLite @charliemarsh-oai
• openai#47317 Enable /import in remote and local daemon sessions @etraut-openai
• openai#47318 Offer explicit recovery for incompatible background servers @etraut-openai
• openai#47319 Match agent status symbols to the selected row style @etraut-openai
• openai#47320 Fix Escape navigation in read-only agent sessions @etraut-openai
• openai#47321 Honor the system clock preference across more TUI timestamps @etraut-openai
• openai#47322 Render TUI Markdown lists with Unicode bullets and checkboxes @etraut-openai
• openai#47323 Persist resume metadata in compaction checkpoints @owenlin0
• openai#47325 Keep report reason tags within Sentry limits @ningyi-oai
• openai#47326 Restrict MCP OAuth authorization endpoints to HTTP(S) @viyatb-oai
• openai#47327 Preserve image generation request IDs on failure @chrisdong-oai
• openai#47329 Standardize TUI menu description wording and punctuation @etraut-openai
• openai#47330 Identify failed SQLite databases in codex doctor output @dkovalenko-oai
• openai#47331 Avoid redundant spacing above the transcript composer @fcoury-oai
• openai#47332 Add GPT-6 Sol and Luna to the model catalog @andrewgu-oai
• openai#47338 Move the activity inspection hint into keyboard shortcut help @fcoury-oai
• openai#47340 Add conditional turn interruption that preserves pending input @pmccrary-oai
• openai#47342 Skip stored title lookups for ephemeral forks @keyz
• openai#47347 Add GPT-6 Sol and Luna to Amazon Bedrock catalogs @celia-oai
• openai#47348 Overlap subagent spawn persistence and clean up cancelled children @owenlin0
• openai#47349 Move plugin recommendations into developer context @adaley-openai
• openai#47350 Reuse cloud skill catalogs across turns until invalidated @hlevy-openai
• openai#47353 Preserve invalid_prompt as a distinct error classification @rennie-openai
• openai#47355 Reserve spacing after recaps at the transcript tail @fcoury-oai
• openai#47358 Color paths and URLs in codex doctor by check status @dkovalenko-oai
• openai#47360 Preserve user prompt URL destinations across terminal widths @fcoury-oai
• openai#47361 Restrict Windows sandbox default object access to the logon session @viyatb-oai
• openai#47362 Bound inbound exec-server requests across client transports @viyatb-oai
• openai#47365 Resume model context from the latest compaction boundary @owenlin0
• openai#47369 Overlap agent metadata reads when resuming a V2 root @vkg-oai
• openai#47375 Add an opt-in preference for the local MXC sandbox @iceweasel-oai
• openai#47377 Add opt-in reasoning status for realtime V3 delegations @guinness-oai
• openai#47380 Route TUI voice controls through the app @etraut-openai
• openai#47381 Keep voice conversations running across TUI thread navigation @etraut-openai
• openai#47382 Show a voice badge in the agents overview @etraut-openai
• openai#47384 Restore composer hints after clearing selection at the transcript bottom @fcoury-oai
• openai#47389 Enforce network policy throughout HTTP and WebSocket requests @acrognale-oai
• openai#47393 Retry transient OpenAI file blob upload failures @mtsui-oai
• openai#47397 Refresh bundled model metadata and instructions @andrewgu-oai
• openai#47399 Respect tmux mouse settings in fullscreen and overlays @fcoury-oai
• openai#47407 Enforce application network policy across app-server requests @acrognale-oai
• openai#47408 Enforce application network policy for AWS auth and telemetry @acrognale-oai
• openai#47410 Honor network policy in remote control and recover remote execution @acrognale-oai
• openai#47411 Apply shared network policy throughout embedded Codex startup @acrognale-oai
• openai#47413 Cache decrypted gateway OAuth secrets per backend @charliemarsh-oai
• openai#47414 Support Shift-click to extend transcript selections @etraut-openai
• openai#47417 Use native scrollback for Terminal.app over SSH in auto mode @fcoury-oai
• openai#47422 Recover unsent question drafts when TUI turns end @imac-oai
• openai#47423 Preserve history search during background draft recovery @imac-oai
• openai#47424 Clear pending async question notifications when turns end @imac-oai
• openai#47428 Apply inherited environment settings at turn boundaries @sayan-oai
• openai#47435 Allow environment selection updates during an active turn @sayan-oai
• openai#47437 Stabilize retry timing and WebSocket tests @anp-oai