Codex CLI

Codex CLI is OpenAI's answer to the agentic coding assistant question — and the answer is fast. Like, noticeably faster than the competition. It's a terminal-based coding agent powered by OpenAI's latest models that takes natural language instructions and turns them into actual code changes in your repo. No GUI needed, no VS Code extension, just your terminal and an API key.

The core model is suggest-then-apply. By default, Codex shows you what it wants to change and waits for your approval. This is actually smart for a default — you see the diff, you approve or reject, you stay in control. But once you trust it, switch to full-auto mode with --full-auto and it just goes. Writes code, runs tests, fixes errors, commits. The whole loop.

The .codex/instructions.md file is where you make it smart about your project. This is your chance to tell it about your stack, your conventions, your testing setup, your deployment process. Codex reads this before every task, so it's not just guessing about how your project works — it actually knows. Think of it as the onboarding doc you wish every new hire would actually read.

Where Codex really shines is batch operations. Need to update 50 API endpoints to use a new auth pattern? Add error handling to every controller? Migrate a config format across an entire codebase? Codex handles the "do this repetitive thing 100 times" class of problems better than almost anything else. It's methodical, it doesn't get bored, and it doesn't miss edge cases on file 47 because it's tired.

Setup is clean: set your OPENAI_API_KEY, install with npm, run codex in your project directory. That's it. No config files to create, no workspace setup, no onboarding wizard. First run takes maybe 30 seconds from install to writing code. The simplicity is actually a feature — less ceremony means more actual work getting done.

The sandbox execution is a nice safety feature. Codex can run commands in a sandboxed environment so it can test its changes without touching your actual system. This means it can run your test suite, check build output, and validate its work without risk. For anything beyond trivial changes, this is the mode you want — it catches problems before they become your problems.

Model selection matters here. Codex supports OpenAI's full model lineup. Use the fast models for quick fixes and small tasks. Use the reasoning models for complex architecture decisions and debugging gnarly bugs. The difference in output quality between models is significant — don't cheap out on the hard problems.

The session history is stored locally in ~/.codex/sessions/. Every conversation, every change, every decision — it's all there. You can resume sessions, review what happened, and trace back why a particular change was made. This is way better than the "black box" approach where you just get a PR and have no idea what the AI was thinking.

Integration with OpenAI's ecosystem is tight. If you're already using OpenAI for other things — embeddings, fine-tuned models, batch processing — Codex fits naturally into that stack. The API key you're already using probably works. No new accounts, no new billing setups, no new vendor relationships.

Who's it for? Engineers who want speed and prefer OpenAI's models. Teams already in the OpenAI ecosystem who want an agentic tool that works with their existing setup. People doing bulk operations across large codebases. If you want the fastest path from "I need this changed" to "it's changed, tested, and committed" — Codex CLI is that path.

## New Features - Save reusable prompts and slash commands in ten numbered `/presets` slots, with a keyboard menu for quick access. (#54380, #54381) - Set model, reasoning effort, speed, and permission profiles directly with inline `/model` and `/permissions` arguments. (#54378, #54383) - Preview viewed and generated images in the fullscreen transcript on terminals supporting Kitty or Sixel graphics. (#54216) - Authenticate enterprise MCP connections using eligible Codex account grants or dedicated sign-in, including OAuth clients requiring secrets. (#55334, #55859, #56001) - Mask credentials during supported remote Bash commands and Zsh snapshot sessions, including credentials from custom providers. (#54309, #54432, #56049) - The Amazon Bedrock setup wizard is now enabled by default, with API key entry grouped alongside other AWS sign-in methods when credentials are detected. (#56033, #56023) ## Bug Fixes - Failed compaction no longer leaves partial summaries in conversation history, and background compaction failures fall back without panicking. (#53932, #55581) - Resuming sessions now finds compressed histories and restores saved thread settings instead of replacing them with global defaults. (#54008, #55923) - Windows HTTPS and WebSocket connections now handle proxy authentication challenges using the signed-in user’s Negotiate or NTLM credentials. (#55994, #56000) - CLI configuration overrides are preserved when `-c` or `--config` appears at multiple subcommand levels. (#56022) - Queued prompts now wait for pending permission selections to resolve before continuing. (#55682) - Hardened sandbox checks for case-variant metadata paths on macOS and replaceable symlinks used to locate executables. (#54273, #55384) ## Chores - Reduced disk usage by pruning unused installations, cleaning abandoned marketplace checkouts, and bounding daemon logs. (#53827, #55149, #55689, #55690, #55856) - Updated the embedded JavaScript runtime with safety checks for optimized `Array.sort` operations. (#54519) ## Changelog Full Changelog: https://github.com/openai/codex/compare/rust-v0.163.0...rust-v0.164.0 - #53725 Render Markdown web links as labels in Rex @bc-openai - #53728 Preserve the remote cwd filter in the in-app resume picker @fcoury-oai - #53732 Reuse compiled search patterns across compressed rollouts @ethernetdan-openai - #53733 Default Guardian v2 async classification to conversation mode @felixxia-oai - #53736 Acquire and cache opt-in executor capability inventories via V2 @TAFOYA-OAI - #53745 Remove `modelSlug` from the latest-model resolver output @kkahadze-oai - #53747 Prefer remote Visualize installs over the bundled plugin @philipp-oai - #53751 Apply feature overrides when restarting the app-server daemon @jif-oai - #53753 Extend copy-on-select to composer text @fcoury-oai - #53754 Support Command/Super shortcuts in the TUI keymap @fcoury-oai - #53763 Enrich V2 capability discovery with resolved plugin and skill metadata @TAFOYA-OAI - #53780 Remove redundant Codex manual cache directory check @kkahadze-oai - #53781 Remove the product-gating TODO from `SkillPolicy` @kkahadze-oai - #53784 Forward client voice session IDs in realtime requests @chrisdong-oai - #53790 Fix cancellation of macOS user-verification prompts @bz-openai - #53794 Retain rollout migration locks across worker cancellation @jif-oai - #53795 Use the bundled model catalog in the MCP startup invalidation test @OLI-OAI - #53797 Refresh Guardian reviews when review context changes @jif-oai - #53800 Wire background compaction into the turn lifecycle @rka-oai - #53804 Keep the Agent command center visible during archive and delete @etraut-openai - #53805 Add safe diagnostics to workspace routing configuration errors @etraut-openai - #53811 Support daemon installation from Windows WinGet packages @etraut-openai - #53816 Reject Guardian review threads during interactive CLI resume @dylan-hurd-oai - #53827 Prune unused app-server daemon releases safely @etraut-openai - #53831 Track content provenance through prompt assembly and compaction @tz-openai - #53845 Fix daemon updater socket binding and bootstrap readiness on Unix @etraut-openai - #53853 Reuse staged file digests when preparing daemon installs @charliemarsh-oai - #53859 Record bucketed terminal sizes in TUI telemetry @imac-oai - #53872 Refresh model instructions in incremental-tool context windows @rhan-oai - #53879 Preserve running turn recovery across thread settings updates @bc-openai - #53895 Resume unloaded subagent threads when selected in the TUI @etraut-openai - #53897 Isolate daemon package tests to reduce disk contention @etraut-openai - #53908 Keep the agent dashboard browsable during task removal @etraut-openai - #53932 Preserve conversation history when compaction fails @jif-oai - #53940 Bound rollout input for memory extraction @jif-oai - #53947 Stop rollout reads when the calling task is cancelled @jif-oai - #53963 Test memory extraction budgets with excluded items and ciphertext @jif-oai - #53969 Avoid reconnect races in the daemon restart test @jif-oai - #53987 Skip service backoff for stale Guardian authorization retries @jif-oai - #53990 Keep host MCP and plugin configuration local with remote environments @jif-oai - #53991 Report the reason for ChatGPT token refresh requests @tejsudha-openai - #53996 Bump `rustls` to `0.23.46` @jif-oai - #53997 Filter rollout filenames before checking compressed siblings @bbrown-oai - #54003 Include compressed rollouts in doctor storage statistics @jif-oai - #54008 Fix session lookup and resume after rollout compression @jif-oai - #54027 Preserve Windows startup variables and MXC child environment defaults @iceweasel-oai - #54032 Bound concurrent Git diffs for untracked files @yuzhu-oai - #54033 Extend memory extension resource retention to 14 days @josh-oai - #54039 Return async Guardian approvals before review cleanup finishes @jif-oai - #54044 Keep fork lineage ahead of paginated transcript history @fcoury-oai - #54046 Track approval cancellation sources and Guardian context changes @vkg-oai - #54047 Restore elapsed time for replayed in-progress turns @etraut-openai - #54048 Defer folder trust checks until starting a task from agents overview @etraut-openai - #54050 Scope TUI skill refreshes and workspace commands to explicit directories @etraut-openai - #54053 Add Responses API client metadata patches for turn start and steer @marksteinbrick-oai - #54056 Coalesce overlapping native TLS root certificate loads @jif-oai - #54057 Distinguish unsupported evidence in Decisions comparison metrics @swastik-oai - #54058 Forward turn metadata to image generation and edit requests @duc-openai - #54059 Support store-owned preparation when resuming threads @owenlin0 - #54060 Limit unarchive scans to the requested thread's archived rollouts @bz-openai - #54069 Fix Windows environment handling for remote MCP servers @iceweasel-oai - #54070 Expose gateway authentication as a reusable OAuth session API @smurching-oai - #54076 Avoid false URL-launch detections in literal PowerShell commands @jif-oai - #54078 Update `gix` to 0.89.0 and `gix-url` to 0.40.0 @jif-oai - #54079 Retire rejected MCP OAuth refresh tokens @mzeng-openai - #54080 Extract TUI user-turn submission into a dedicated module @bc-openai - #54081 Centralize vertical navigation in TUI list views @bc-openai - #54082 Stabilize the Guardian network review timeout test @felixxia-oai - #54087 Upgrade Bazel's `rules_python` to 2.4.2 @dkovalenko-oai - #54094 Preserve specific MCP failure codes in metrics and spans @TAFOYA-OAI - #54096 Add Bazel targets for on-demand schemas and Python SDK generation @dkovalenko-oai - #54097 Show Daybreak notices after the session header @charliemarsh-oai - #54098 Keep environment tools stable across readiness changes @aibrahim-oai - #54099 Wait for readiness in the SIGTERM cancellation test @felixxia-oai - #54113 Embed generated app-server schema bundles in Bazel builds @dkovalenko-oai - #54114 Report terminal completion states and blocking reasons @cconger - #54119 Allow stdin review under active grants with sandbox prompts disabled @tepman-oai - #54133 Record per-tool incremental update and description size metrics @akira-oai - #54135 Allow large file writes over direct exec-server WebSockets @ostepanian - #54151 Generate keymap descriptors and binding accessors from one catalog @bc-openai - #54169 Reload the model catalog after account changes @smurching-oai - #54170 Preserve ChatGPT authentication for Gov MCP servers @lsingh-oai - #54175 Retire worktree and voice feature flags @bc-openai - #54216 Show image previews in the fullscreen transcript @etraut-openai - #54217 Apply syntax theme colors to Markdown prose in the TUI @etraut-openai - #54250 Add regression tests for case-variant metadata protection @jif-oai - #54257 Preserve recent speech when trimming realtime transcripts @jif-oai - #54263 Require fresh Guardian approval for escalated tool retries @jif-oai - #54264 Track Guardian transcript admission across failed reviews @jif-oai - #54265 Add regression coverage for Guardian sandbox retry approvals @jif-oai - #54273 Protect metadata writes through macOS case aliases @jif-oai - #54278 Share Guardian image deduplication and checkpoint restoration @felixxia-oai - #54281 Attribute compaction summaries to the assistant in Guardian transcripts @jif-oai - #54309 Enable credential masking for remote exec-server commands @viyatb-oai - #54315 Keep TUI Markdown colors readable against their backgrounds @fcoury-oai - #54316 Centralize context fragment kind constants @ioanbudea-OAI - #54346 Add scoped executor credential auth for remote exec-server @viyatb-oai - #54351 Use the turn index for turn-scoped item pagination @bz-openai - #54363 Filter uninspectable network roots from implicit MXC read grants @iceweasel-oai - #54365 Prevent duplicate assistant messages around interleaved commands @fcoury-oai - #54369 Retry SQLite operations on transient writer contention @willwang-openai - #54378 Support inline model, reasoning, and speed selection in `/model` @fcoury-oai - #54379 Add configuration support for TUI input presets @fcoury-oai - #54380 Add `/presets` for reusable prompts and slash commands @fcoury-oai - #54381 Add a keyboard shortcut menu for presets @fcoury-oai - #54383 Support inline permission profile selection with `/permissions` @fcoury-oai - #54385 Allow saving current permission settings as presets @fcoury-oai - #54387 Add session-only arguments to `/title` and `/statusline` @fcoury-oai - #54388 Add quick dial presets for terminal title and status line layouts @fcoury-oai - #54389 Preserve existing SQLite journal modes on startup @dkovalenko-oai - #54398 Store generated image items as attachment references @kchainani-oai - #54401 Run managed daemon processes at background priority on macOS @bc-openai - #54414 Allow root turn suspension with live descendants @miz-openai - #54417 Skip SQLite page reclamation on Linux NFS mounts @dkovalenko-oai - #54419 Support line breaks in Mermaid flowchart node labels @etraut-openai - #54420 Record local image inputs on completed image-generation items @mjkaufer-openai - #54421 Treat explicit connections to the verified default daemon as local @etraut-openai - #54424 Add remote plugin metadata to plugin configuration @mzeng-openai - #54432 Support remote credential masking with Zsh shell snapshots @viyatb-oai - #54433 Use neutral labels for nonzero exits in compact TUI output @bc-openai - #54434 Add rotating stdin credentials for remote exec-server @viyatb-oai - #54445 Remove obsolete protected-path monitoring from the Linux sandbox @viyatb-oai - #54475 Use a synchronous mutex for remote metadata request tracking @anp-oai - #54505 Add streaming file replacement to filesystem abstractions @anp-oai - #54519 Upgrade V8 to 152.2.0 and backport Array.sort safety guards @seanh-oai - #54589 Preserve evidence message structure in Guardian Decisions requests @swastik-oai - #54590 Capture sender context for `codex_app.create_thread` deliveries @felixxia-oai - #54619 Bind plugin hook approval to plugin contents @jif-oai - #54623 Resolve stable releases dynamically for exec-server compatibility tests @felixxia-oai - #54637 Preserve disconnect reasons in Noise handshake errors @virajrai-oai - #54639 Stream rollout records during thread history materialization @chess-oai - #54641 Retry npm release publishing and match macOS verification runners @jif-oai - #54647 Make remote-control conflict retry tests deterministic @jif-oai - #54653 Prefer resource-specific scopes during MCP OAuth discovery @zamoshchin-openai - #54654 Allow dynamic tool calls to execute in parallel @sdcoffey - #54656 Honor MCP tool-list changes within a turn @jif-oai - #54664 Distinguish retryable thread-store errors from unknown write outcomes @wiltzius-openai - #54672 Use shared Bazel CI preparation for lint and V8 jobs @anp-oai - #54674 Add an opt-in tool for copying files between environments @anp-oai - #54679 Distinguish MCP approval cancellation reasons @vkg-oai - #54681 Keep TUI session selections bound to their backend @bc-openai - #54686 Preserve MCP tool-list metadata in server status @victor-openai - #54708 Honor home deletion preferences during Windows uninstall @chess-oai - #54710 Move V8's ICU Bazel rules into the workspace @seanh-oai - #54716 Keep prereleases in draft until release assets finish uploading @ioanbudea-OAI - #54724 Preserve prereleased events when creating GitHub prereleases @ioanbudea-OAI - #54737 Enable BuildBuddy for voice release builds @anp-oai - #54739 Close arg0 lock files before removing temporary directories @dkovalenko-oai - #54741 Gate the file-copy extension behind `multi_env_copy_file` @anp-oai - #54742 Render slash command prefixes as composer prompt markers @bc-openai - #54754 Reduce memory overhead in tool JSON schemas @charliemarsh-oai - #54761 Upgrade `zip` to 9.0.0 and update plugin archive extraction @anp-oai - #54763 Stop applying macOS background priority to managed daemons @bc-openai - #54766 Defer unknown TUI config fields to the shared strict-config policy @etraut-openai - #54776 Support typed content in contextual prompt fragments @pakrym-oai - #54930 Add negotiated zstd compression to app-server WebSockets @thorvald-oai - #54945 Share additional tool declarations across response item clones @charliemarsh-oai - #54946 Preserve model catalog caches across header credential rotation @hlevy-openai - #54949 Gate app-server WebSocket compression behind a feature flag @thorvald-oai - #54961 Allow agent messages to wake idle root sessions behind a feature flag @alishobeiri-oai - #55144 Add observation-only model request lifecycle hooks @vivi - #55149 Prune unused standalone releases after successful installation @etraut-openai - #55226 Bucket admitted Guardian request metrics by input size @ioanbudea-OAI - #55227 Allow plus signs in metric tags @swastik-oai - #55234 Record Guardian sampling retries, attempts, and delivery outcomes @ioanbudea-OAI - #55237 Add Guardian review routing and cached score metrics @ioanbudea-OAI - #55244 Batch thread history materialization into SQLite @jif-oai - #55252 Preserve rollout access metadata across compression and restoration @jif-oai - #55255 Unload idle ephemeral composer prediction threads without delay @jif-oai - #55261 Share dynamic tool specifications across turn snapshots @charliemarsh-oai - #55262 Stop memory dual-write when v2 is selected @jif-oai - #55263 Start dormant MCP servers for hook calls @charliemarsh-oai - #55265 Avoid cloning the full catalog for model metadata lookups @charliemarsh-oai - #55270 Honor per-thread unload delays in app-server @jif-oai - #55274 Stabilize daemon reconnect and Rosetta package startup tests @jif-oai - #55278 Fall back when macOS rejects high stdio descriptors @chrishayduk-oai - #55283 Avoid cloning events in the app-server thread listener @charliemarsh-oai - #55287 Track MCP enterprise auth cancellation across config reloads @jif-oai - #55288 Recheck EMA request authority at HTTP boundaries @jif-oai - #55293 Keep only the latest adjacent reasoning effort update in requests @felixxia-oai - #55296 Encapsulate Responses API proxy authentication @ee-oai - #55297 Add GitHub Actions OIDC token acquisition to the responses API proxy @ee-oai - #55298 Add renewable federated auth primitives to the Responses API proxy @ee-oai - #55299 Refresh rejected proxy credentials and back off failed renewals @ee-oai - #55300 Add one-shot app-server startup configuration discovery @acrognale-oai - #55301 Reconcile V2 executor plugins with cloud catalogs @TAFOYA-OAI - #55302 Stop the closest remote SHA search at zero distance @jif-oai - #55303 Route skill reads by alias source @TAFOYA-OAI - #55304 Reject TUI events from a different backend @bc-openai - #55320 Enable GitHub Actions OIDC authentication in the Responses API proxy @ee-oai - #55322 Integrate discovery V2 with step-scoped skill catalogs @TAFOYA-OAI - #55324 Remove thread title generation spinners from the TUI @cconger - #55327 Add configured thread environment inheritance for subagents @jif-oai - #55330 Expose skill installations and stabilize package aliases @TAFOYA-OAI - #55331 Use step environment snapshots for subagent inheritance @jif-oai - #55334 Enable enterprise-managed authorization for MCP connections @jif-oai - #55339 Add experimental `thread/sample` requests to the app server @liame-oai - #55342 Consolidate cached and fetched Codex manual handling @kkahadze-oai - #55347 Add Guardian retry metrics and preserve turn attribution @ioanbudea-OAI - #55351 Reduce path storage in the file-search index @charliemarsh-oai - #55355 Remove the long-thread accuracy warning after compaction @asohrabi-oai - #55357 Prepare skill catalogs before building step tools @TAFOYA-OAI - #55358 Add a shared cleanup lock guard for local log directories @dkovalenko-oai - #55365 Import only the continuation of backgrounded Claude sessions @jgrodman-arch - #55366 Support catalog templates for approved command prefixes @dylan-hurd-oai - #55368 Enable incremental tools by default for Responses Lite @akira-oai - #55372 Fix recall of command-line initial prompts in the TUI @joesavona - #55375 Add configured thread environment inheritance for subagents @jif-oai - #55378 Migrate the Windows MXC sandbox to `mxc-sdk` @johnl-oai - #55379 Handle response body failures in OpenAI docs downloads @kkahadze-oai - #55380 Preserve thread metadata before rollout files are materialized @jif-oai - #55381 Manage TUI backend sessions independently @bc-openai - #55384 Reject replaceable symlinks in pre-sandbox executable lookup @keeley-openai - #55387 Add support for Academic Researchers plans @etraut-openai - #55388 Mark `transcript_v2` as removed @etraut-openai - #55391 Deduplicate native agent messages in Guardian review context @jif-oai - #55394 Use thread connections for enterprise MCP status listings @jif-oai - #55395 Roll back interrupted Guardian reviews to preserve evidence @jif-oai - #55396 Prefetch sibling rollout paths when reading history pages @bz-openai - #55397 Record OAuth provenance and login identity for account tokens @jif-oai - #55398 Add wire types for runtime agent registration @cooper-oai - #55401 Defer environment disconnect notifications until recovery fails @richardopenai - #55402 Add ephemeral runtime registration and signed assertions @cooper-oai - #55406 Trace harness-side Noise handshake failures @alishobeiri-oai - #55407 Clarify API key login and status messages @fcoury-oai - #55409 Retry TUI startup after a server shutdown refusal @fcoury-oai - #55410 Track credential sources separately from auth owners @cooper-oai - #55411 Render async question prompts as Markdown in the TUI @bc-openai - #55413 Share runtime authentication across requests and credential refreshes @cooper-oai - #55441 Add cancellation for experimental thread samples @liame-oai - #55442 Check daemon feature overrides and prompt for mismatch recovery @etraut-openai - #55456 Support Windows registry overrides for `requirements.toml` @johnl-oai - #55463 Add optional MCP disablement for prediction forks and samples @liame-oai - #55498 Add correlated diagnostics for Noise relay handshakes @virajrai-oai - #55524 Make the startup migration contention test deterministic @jif-oai - #55533 Separate Guardian conversation preparation from execution @swastik-oai - #55545 Improve diagnostics for compressed rollout read failures @jif-oai - #55546 Add scoped leases for persisted account OAuth grants @jif-oai - #55547 Ignore the basic-toml advisory in dependency audits @jif-oai - #55548 Preserve parent history by default in full agent forks @jif-oai - #55581 Avoid panics after background compaction fails before a checkpoint @jif-oai - #55587 Allow callers to control prompt cache keys independently of session IDs @rm-openai - #55591 Measure Guardian approval blocking time per turn @ioanbudea-OAI - #55593 Reject unpackaged Windows sandbox provisioning clients @chess-oai - #55596 Rank `$` mentions by recent selection frequency @bc-openai - #55597 Extract shared TUI user prompt helpers from IDE context @bc-openai - #55607 Propagate authentication header failures before sending requests @cooper-oai - #55616 Update `rustls-webpki` and `webpki-roots` lockfile versions @jif-oai - #55620 Move Linux sandbox unit tests into separate files @bolinfest - #55623 Refactor Guardian v2 Decisions snapshot comparison @swastik-oai - #55627 Reserve a sibling rollout candidate slot for the first ancestor @bz-openai - #55631 Add frozen cloud skill snapshots to the app server @fc-oai - #55633 Treat code-mode collaboration messages as plaintext @hlevy-openai - #55635 Enable Decisions comparisons for Guardian V2 conversation mode @swastik-oai - #55636 Add runtime assertion authentication modes and an auth provider @cooper-oai - #55644 Make pending injections interrupt agent waits and sleep @jif-oai - #55645 Fix in-process app-server shutdown with full queues @jif-oai - #55646 Guard auth persistence and reloads against concurrent login changes @jif-oai - #55652 Search compressed rollouts concurrently @ethernetdan-openai - #55653 Track code-mode cell lifecycle and result buffering time @euroelessar - #55657 Add lifecycle hooks for inter-agent messages @euroelessar - #55666 Omit reasoning and compaction from Decisions snapshot comparisons @swastik-oai - #55667 Enable app routing for remote curated Agent Plugins @tsarlandie-oai - #55677 Upgrade the Rust toolchain to 1.99.1 @dkovalenko-oai - #55680 Recheck account login ownership for enterprise MCP @jif-oai - #55682 Fix TUI permission selection and session state handling @etraut-openai - #55685 Add account credential support to EMA token exchange helpers @jif-oai - #55689 Clean up failed and abandoned marketplace staging directories @dkovalenko-oai - #55690 Reclaim aged legacy marketplace staging directories @dkovalenko-oai - #55691 Refine enterprise MCP auth errors and request authority checks @jif-oai - #55692 Tag tool registration metrics with effective incremental tools mode @akira-oai - #55693 Use hermetic dsymutil for macOS Bazel builds @dkovalenko-oai - #55696 Preserve enterprise MCP authorization across runtime refreshes @jif-oai - #55698 Reuse Codex account grants for enterprise MCP authentication @jif-oai - #55701 Enable MCP OAuth refresh coordination by default @mzeng-openai - #55705 Centralize cloud skill cache state under the catalog lock @TAFOYA-OAI - #55710 Resolve MCP backend identity from runtime events in rollout traces @cassirer-openai - #55719 Allow unified exec to fall back when credential masking is unavailable @viyatb-oai - #55723 Avoid redundant clones when sending app-server notifications @charliemarsh-oai - #55724 Share prepared MCP tool definitions across sessions @charliemarsh-oai - #55725 Preserve code cell response metadata in rollout traces @cassirer-openai - #55733 Box session metadata and compaction payloads in `RolloutItem` @charliemarsh-oai - #55804 Add a critical-path utility for exclusive duration attribution @taohong339 - #55846 Measure Guardian review latency once with a monotonic clock @ioanbudea-OAI - #55847 Bound retained turn analytics and clean up closed threads @jif-oai - #55848 Pause relay test time while inspecting handshake diagnostics @jif-oai - #55856 Bound managed app-server and updater stderr logging @jif-oai - #55859 Fix enterprise MCP credential selection and reauthentication @jif-oai - #55883 Expand enterprise MCP authorization lifecycle test coverage @jif-oai - #55913 Skip conversation scoring superseded by newer scores and history @jif-oai - #55923 Preserve thread model settings across resume and revert @etraut-openai - #55940 Add local diagnostic ZIP exports to the app server @viktor-oai - #55946 Bound remote installed-plugin inventory pagination @mzeng-openai - #55954 Separate retained instruction collection from budget policy @ioanbudea-OAI - #55957 Refresh cloud plugin and skill catalogs after mid-turn auth changes @TAFOYA-OAI - #55958 Preserve provider model catalogs when refreshes fail @abhinav-oai - #55965 Add managed Desktop workspace storage requirements @acrognale-oai - #55968 Back off failed cloud skill and plugin catalog refreshes @TAFOYA-OAI - #55969 Drain WebSocket responses on mailbox preemption @jif-oai - #55970 Extract shared message-board code into `codex-message-board` @felixxia-oai - #55971 Add storage namespaces for local message boards @felixxia-oai - #55973 Allow consumers to select message-board tools @felixxia-oai - #55975 Add an experimental command to migrate NFS databases to DELETE mode @dkovalenko-oai - #55978 Route SQLite logs to an explicitly configured `log_dir` @dkovalenko-oai - #55987 Refresh remote exec-server credentials after unauthorized responses @apanasenko-oai - #55991 Avoid double-counting retained reasoning in token usage @dkovalenko-oai - #55993 Preserve per-execution trace context in code-mode stdio callbacks @bschoepke-openai - #55994 Add a Windows proxy authentication tunnel helper @efrazer-oai - #56000 Retry proxy CONNECT challenges with Windows authentication @efrazer-oai - #56001 Support confidential OAuth clients for enterprise-managed MCP auth @jif-oai - #56003 Scope MITM credential substitution to trusted origins and placeholders @viyatb-oai - #56006 Assert shell tool overlap directly in the parallelism test @goodman-oai - #56007 Pause Tokio time in the reused Guardian session test @goodman-oai - #56008 Add a critical-path timing recorder for direct tool calls @taohong339 - #56011 Keep rollout trace terminal sessions distinct when process IDs are reused @cassirer-openai - #56018 Avoid nested CLI startup in the MCP OAuth persistence test @goodman-oai - #56020 Allow thread sampling after failed or interrupted turns @liame-oai - #56021 Keep slash command prefixes in editable composer text @bc-openai - #56022 Preserve config overrides across CLI subcommand boundaries @jmickey-oai - #56023 Group Bedrock API key entry with other AWS sign-in methods @alexsong-oai - #56024 Limit decoded HTTP/3 response headers in the TCP tunnel @richardopenai - #56031 Support client Responses API metadata in `thread/sample` @liame-oai - #56032 Fix entrypoint lookup in package debug-symbol smoke tests @dkovalenko-oai - #56033 Enable the Bedrock setup wizard by default @alexsong-oai - #56036 Add request-scoped OAuth credentials and session lifecycle APIs @smurching-oai - #56037 Add a channel-backed cloud app-server client @bc-openai - #56044 Enable executor cleanup hooks with V2 capability discovery @TAFOYA-OAI - #56049 Support configured credential providers in remote execution @viyatb-oai - #56055 Require managed configuration from the configured server @smurching-oai

View on GitHub ↗

.codex/instructions.md is your foundation

Write it once. Codex reads it every run. Put your stack, conventions, testing patterns, and rules here. It's the difference between a tool that guesses and one that knows your codebase.

exec for one-shot CI runs

Non-interactive mode. Perfect for automation, CI pipelines, and scripted tasks.

codex exec "run tests and fix any failing ones"

Approval mode determines autonomy

Three modes: suggest (show changes, ask), auto (apply without asking), full-auto (run commands too).

codex "task" --approval-mode auto

Commit before big tasks

Codex reads your git history as part of its context. A clean commit before a big task also gives you a clear rollback point.

Feed it the whole repo

Don't point Codex at one file. It builds a mental model of your whole codebase and makes better decisions when it has full context.

--quiet for clean script output

Suppresses the "thinking" output for cleaner integration into shell pipelines and CI logs.

codex --quiet exec "task"

🤖

Built-in sub-agent support

For large tasks, Codex can spawn parallel worker agents automatically. It handles the orchestration — you don't need to break the task up manually. Just describe the full scope and it figures out the parallelism.

🔗

Give it a URL — it fetches and reads

You can reference a URL in your task description and Codex will fetch and read it as context. Useful for "implement this API according to the spec at https://..."

🔄

Swap models with --model

Use a different model for different tasks without changing config. Fast model for exploration, smarter model for production code.

codex --model gpt-4o-mini "quick refactor of this function"

☠️

No instructions.md

Codex guesses your stack and conventions from scratch every run. Everything is slower and less accurate without it.

Create .codex/instructions.md before your first real task.

☠️

Using full-auto on unfamiliar code

full-auto runs shell commands without confirmation. In code you don't know well, this can delete files or run destructive migrations.

Use suggest on code you don't own. full-auto only for trusted, well-tested repos.

☠️

Skipping git commit before big tasks

If you don't like Codex's output, you need a clean commit to roll back to. Working with a dirty git state gives you no safety net.

git add -A && git commit -m "checkpoint" before every big Codex task.

CLI Flags

--full-auto Convenience alias for `--sandbox workspace-write --ask-for-approval on-request`. Best default for hands-off local edits inside the repo.

--dangerously-bypass-approvals-and-sandbox Disables both sandboxing and approval prompts. Highest-risk mode.

--sandbox Sets sandbox mode: `read-only`, `workspace-write`, or `danger-full-access`.

--ask-for-approval Controls when Codex asks before executing commands: `untrusted`, `on-request`, or `never` (plus deprecated `on-failure`).

--cd Sets the working root without manually changing directories.

--add-dir Adds extra writable roots alongside the primary workspace.

--skip-git-repo-check Allows `codex exec` to run outside a Git repository.

--json Streams newline-delimited JSON events instead of human-formatted output in `exec` mode.

--output-last-message Writes the final assistant message to a file in `exec` mode.

--output-schema Constrains the final response shape using a JSON Schema file.

--ephemeral Runs without persisting session files to disk.

--model Selects the model for the run or session.

--profile Loads a named config profile from `config.toml`.

--search Enables live web search instead of cached web results.

--image Attaches one or more images to the initial prompt.

--config One-off override for any config key; values are parsed as TOML.

Daily Workflows

Quick one-shot fix codex exec --full-auto "fix the failing test, run the narrowest relevant test, and summarize what changed"

Interactive debugging in a real repo codex "investigate why login sometimes fails after refresh; explain root cause before editing"

Plan first, then implement codex "propose a step-by-step plan to add rate limiting to the API; do not edit yet"

Resume yesterday's thread codex resume --last

Continue a past non-interactive run codex exec resume --last "implement the fix you proposed and run focused validation"

Review your local diff before commit codex exec review "review the current changes for correctness, regressions, and missing tests"

Patch across multiple directories safely codex --cd apps/frontend --add-dir ../backend --add-dir ../shared "implement the API + UI change end to end"

Headless/remote login setup codex login --device-auth

Machine-readable CI run codex exec --json -o /tmp/codex-last.txt "summarize the CI failure and propose the smallest fix"

Launch a Codex Cloud task from the terminal codex cloud exec --env ENV_ID "implement the feature described in the issue" codex cloud exec --env ENV_ID --attempts 3 "fix the bug with three candidate solutions"

⚡ Power Moves

Parallel worktree army git worktree add ../repo-bug-1 -b bug-1 && git worktree add ../repo-bug-2 -b bug-2 && (cd ../repo-bug-1 && codex exec --full-auto "fix bug A and run focused tests") & (cd ../repo-bug-2 && codex exec --full-auto "fix bug B and run focused tests") & wait

Plan → implement → review loop codex "plan the migration" # then /review after changes

Scoped autonomy instead of YOLO codex --cd services/api --add-dir ../shared "update the contract and consumers"

Codify your workflow in AGENTS.md codex # in a repo with AGENTS.md describing conventions and validation commands

Use `/compact` before the thread gets dumb /compact

JSON schema output for structured pipelines codex exec --output-schema ./result.schema.json "analyze the repo and return structured risks"

Multi-agent for bounded parallel work codex # ask: 'Spawn one agent per review point, wait for all, summarize' # Custom agent: .codex/agents/reviewer.toml with name/description/developer_instructions

Device auth + copied auth cache for remote boxes codex login --device-auth

Codex CLI

Power Tips

Quick Wins

Hidden Features

Common Mistakes

Command Reference