86 Commits

• rollout: persist turn permission profiles (#18281)

  ## Why
  
  Resume and reconstruction need to preserve the permissions that were
  active for each user turn. If rollouts only keep legacy sandbox fields,
  replay cannot faithfully represent profile-shaped overrides introduced
  earlier in the stack.
  
  ## What changed
  
  This records `permission_profile` on user-turn rollout events,
  reconstructs it through history/state extraction, and updates rollout
  reconstruction and related fixtures to keep the field explicit.
  
  ## Verification
  
  - `cargo test -p codex-core --test all permissions_messages --
  --nocapture`
  - `cargo test -p codex-core --test all request_permissions --
  --nocapture`
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  ---
  [//]: # (BEGIN SAPLING FOOTER)
  Stack created with [Sapling](https://sapling-scm.com). Best reviewed
  with [ReviewStack](https://reviewstack.dev/openai/codex/pull/18281).
  * #18288
  * #18287
  * #18286
  * #18285
  * #18284
  * #18283
  * #18282
  * __->__ #18281

  Michael Bolin · 2026-04-22 17:00:29 -07:00
  6ca038bbd1
• Split DeveloperInstructions into individual fragments. (#18813)

  Split DeveloperInstructions into individual fragments.

  pakrym-oai · 2026-04-21 10:22:36 -07:00
  2a226096f6
• Organize context fragments (#18794)

  Organize context fragments under `core/context`. Implement same trait on
  all of them.

  pakrym-oai · 2026-04-20 22:39:17 -07:00
  4c2e730488
• Update image outputs to default to high detail (#18386)

  Do not assume the default `detail`.

  pakrym-oai · 2026-04-18 11:01:12 -07:00
  53b1570367
• Move codex module under session (#18249)

  ## Summary
  - rename the core codex module root to session/mod.rs without using
  #[path]
  - move the codex module directory and tests under core/src/session
  - remove session/mod.rs reexports so call sites use explicit child
  module paths
  
  ## Testing
  - cargo test -p codex-core --lib
  - cargo check -p codex-core --tests
  - just fmt
  - just fix -p codex-core
  - git diff --check

  pakrym-oai · 2026-04-17 16:18:53 +00:00
  71e4c6fa17
• feat(permissions): add glob deny-read policy support (#15979)

  ## Summary
  - adds first-class filesystem policy entries for deny-read glob patterns
  - parses config such as :project_roots { "**/*.env" = "none" } into
  pattern entries
  - enforces deny-read patterns in direct read/list helpers
  - fails closed for sandbox execution until platform backends enforce
  glob patterns in #18096
  - preserves split filesystem policy in turn context only when it cannot
  be reconstructed from legacy sandbox policy
  
  ## Stack
  1. This PR - glob deny-read policy/config/direct-tool support
  2. #18096 - macOS and Linux sandbox enforcement
  3. #17740 - managed deny-read requirements
  
  ## Verification
  - just fmt
  - cargo check -p codex-core -p codex-sandboxing --tests
  
  ---------
  
  Co-authored-by: Codex <noreply@openai.com>

  viyatb-oai · 2026-04-16 10:31:51 -07:00
  6862b9c745
• feat(guardian): send only transcript deltas on guardian followups (#17269)

  ## Description
  
  We reuse a guardian thread for a given user thread when we can. However,
  we had always sent the full transcript history every time we made a
  followup review request to an existing guardian thread.
  
  This is especially bad for long guardian threads since we keep
  re-appending old transcript entries instead of just what has changed.
  The fix is to just send what's new.
  
  **Caveat**: Whenever a thread is compacted or rolled back, we fall back
  to sending the full transcript to guardian again since the thread's
  history has been modified. However in the happy path we get a nice
  optimization.
  
  ## Before
  Initial guardian review sends the full parent transcript:
  
  ```
  The following is the Codex agent history whose request action you are assessing...
  >>> TRANSCRIPT START
  [1] user: Please check the repo visibility and push the docs fix if needed.
  [2] tool gh_repo_view call: {"repo":"openai/codex"}
  [3] tool gh_repo_view result: repo visibility: public
  [4] assistant: The repo is public; I now need approval to push the docs fix.
  >>> TRANSCRIPT END
  The Codex agent has requested the following action:
  >>> APPROVAL REQUEST START
  ...
  >>> APPROVAL REQUEST END
  ```
  
  And a followup to the same guardian thread would send the full
  transcript again (including items 1-4 we already sent):
  ```
  The following is the Codex agent history whose request action you are assessing...
  >>> TRANSCRIPT START
  [1] user: Please check the repo visibility and push the docs fix if needed.
  [2] tool gh_repo_view call: {"repo":"openai/codex"}
  [3] tool gh_repo_view result: repo visibility: public
  [4] assistant: The repo is public; I now need approval to push the docs fix.
  [5] user: Please push the second docs fix too.
  [6] assistant: I need approval for the second docs fix.
  >>> TRANSCRIPT END
  The Codex agent has requested the following action:
  >>> APPROVAL REQUEST START
  ...
  >>> APPROVAL REQUEST END
  ```
  
  ## After
  Initial guardian review sends the full parent transcript (this is
  unchanged):
  
  ```
  The following is the Codex agent history whose request action you are assessing...
  >>> TRANSCRIPT START
  [1] user: Please check the repo visibility and push the docs fix if needed.
  [2] tool gh_repo_view call: {"repo":"openai/codex"}
  [3] tool gh_repo_view result: repo visibility: public
  [4] assistant: The repo is public; I now need approval to push the docs fix.
  >>> TRANSCRIPT END
  The Codex agent has requested the following action:
  >>> APPROVAL REQUEST START
  ...
  >>> APPROVAL REQUEST END
  ```
  
  But a followup now sends:
  ```
  The following is the Codex agent history added since your last approval assessment. Continue the same review conversation...
  >>> TRANSCRIPT DELTA START
  [5] user: Please push the second docs fix too.
  [6] assistant: I need approval for the second docs fix.
  >>> TRANSCRIPT DELTA END
  The Codex agent has requested the following next action:
  >>> APPROVAL REQUEST START
  ...
  >>> APPROVAL REQUEST END
  ```

  Owen Lin · 2026-04-10 07:48:44 -07:00
  88165e179a
• [codex] allow disabling environment context injection (#16745)

  This adds an `include_environment_context` config/profile flag that
  defaults on, and guards both initial injection and later environment
  updates to allow skipping injection of `<environment_context>`.

  Thibault Sottiaux · 2026-04-03 18:06:52 -07:00
  91ca49e53c
• [codex] allow disabling prompt instruction blocks (#16735)

  This PR adds root and profile config switches to omit the generated
  `<permissions instructions>` and `<apps_instructions>` prompt blocks
  while keeping both enabled by default, and it gates both the initial
  developer-context injection and later permissions diff injection so
  turning the permissions block off stays effective across turn-context
  overrides.
  
  Also added a prompt debug tool that can be used as `codex debug
  prompt-input "hello"` and dumps the constructed items list.

  Thibault Sottiaux · 2026-04-03 23:47:56 +00:00
  8d19646861
• chore: clean up argument-comment lint and roll out all-target CI on macOS (#16054)

  ## Why
  
  `argument-comment-lint` was green in CI even though the repo still had
  many uncommented literal arguments. The main gap was target coverage:
  the repo wrapper did not force Cargo to inspect test-only call sites, so
  examples like the `latest_session_lookup_params(true, ...)` tests in
  `codex-rs/tui_app_server/src/lib.rs` never entered the blocking CI path.
  
  This change cleans up the existing backlog, makes the default repo lint
  path cover all Cargo targets, and starts rolling that stricter CI
  enforcement out on the platform where it is currently validated.
  
  ## What changed
  
  - mechanically fixed existing `argument-comment-lint` violations across
  the `codex-rs` workspace, including tests, examples, and benches
  - updated `tools/argument-comment-lint/run-prebuilt-linter.sh` and
  `tools/argument-comment-lint/run.sh` so non-`--fix` runs default to
  `--all-targets` unless the caller explicitly narrows the target set
  - fixed both wrappers so forwarded cargo arguments after `--` are
  preserved with a single separator
  - documented the new default behavior in
  `tools/argument-comment-lint/README.md`
  - updated `rust-ci` so the macOS lint lane keeps the plain wrapper
  invocation and therefore enforces `--all-targets`, while Linux and
  Windows temporarily pass `-- --lib --bins`
  
  That temporary CI split keeps the stricter all-targets check where it is
  already cleaned up, while leaving room to finish the remaining Linux-
  and Windows-specific target-gated cleanup before enabling
  `--all-targets` on those runners. The Linux and Windows failures on the
  intermediate revision were caused by the wrapper forwarding bug, not by
  additional lint findings in those lanes.
  
  ## Validation
  
  - `bash -n tools/argument-comment-lint/run.sh`
  - `bash -n tools/argument-comment-lint/run-prebuilt-linter.sh`
  - shell-level wrapper forwarding check for `-- --lib --bins`
  - shell-level wrapper forwarding check for `-- --tests`
  - `just argument-comment-lint`
  - `cargo test` in `tools/argument-comment-lint`
  - `cargo test -p codex-terminal-detection`
  
  ## Follow-up
  
  - Clean up remaining Linux-only target-gated callsites, then switch the
  Linux lint lane back to the plain wrapper invocation.
  - Clean up remaining Windows-only target-gated callsites, then switch
  the Windows lint lane back to the plain wrapper invocation.

  Michael Bolin · 2026-03-27 19:00:44 -07:00
  61dfe0b86c
• [codex] Defer fork context injection until first turn (#15699)

  ## Summary
  - remove the fork-startup `build_initial_context` injection
  - keep the reconstructed `reference_context_item` as the fork baseline
  until the first real turn
  - update fork-history tests and the request snapshot, and add a
  `TODO(ccunningham)` for remaining nondiffable initial-context inputs
  
  ## Why
  Fork startup was appending current-session initial context immediately
  after reconstructing the parent rollout, then the first real turn could
  emit context updates again. That duplicated model-visible context in the
  child rollout.
  
  ## Impact
  Forked sessions now behave like resume for context seeding: startup
  reconstructs history and preserves the prior baseline, and the first
  real turn handles any current-session context emission.
  
  ---------
  
  Co-authored-by: Codex <noreply@openai.com>

  Charley Cunningham · 2026-03-24 18:34:44 -07:00
  d72fa2a209
• Move string truncation helpers into codex-utils-string (#15572)

  - move the shared byte-based middle truncation logic from `core` into
  `codex-utils-string`
  - keep token-specific truncation in `codex-core` so rollout can reuse
  the shared helper in the next stacked PR
  
  ---------
  
  Co-authored-by: Codex <noreply@openai.com>

  Ahmed Ibrahim · 2026-03-24 15:45:40 -07:00
  062fa7a2bb
• Move git utilities into a dedicated crate (#15564)

  - create `codex-git-utils` and move the shared git helpers into it with
  file moves preserved for diff readability
  - move the `GitInfo` helpers out of `core` so stacked rollout work can
  depend on the shared crate without carrying its own git info module
  
  ---------
  
  Co-authored-by: Ahmed Ibrahim <219906144+aibrahim-oai@users.noreply.github.com>
  Co-authored-by: Codex <noreply@openai.com>

  Ahmed Ibrahim · 2026-03-24 13:26:23 -07:00
  0f957a93cd
• Trim pre-turn context updates during rollback (#15577)

  ## Summary
  - trim contiguous developer/contextual-user pre-turn updates when
  rollback cuts back to a user turn
  - add a focused history regression test for the trim behavior
  - update the rollback request-boundary snapshots to show the fixed
  non-duplicating context shape
  
  ---------
  
  Co-authored-by: Codex <noreply@openai.com>

  Charley Cunningham · 2026-03-24 12:43:53 -07:00
  2d61357c76
• feat: communication pattern v2 (#15647)

  See internal communication

  jif-oai · 2026-03-24 18:45:49 +00:00
  773fbf56a4
• feat: disable notifier v2 and start turn on agent interaction (#15624)

  Make the inter-agent communication start a turn
  
  As part of this, we disable the v2 notifier to prevent some odd
  behaviour where the agent restart working while you're talking to it for
  example

  jif-oai · 2026-03-24 17:01:24 +00:00
  b51d5f18c7
• feat: use serde to differenciate inter agent communication (#15560)

  Use `serde` to encode the inter agent communication to an assistant
  message and use the decode to see if this is such a message
  
  Note: this assume serde on small pattern is fast enough

  jif-oai · 2026-03-23 22:09:55 +00:00
  191fd9fd16
• feat: new op type for sub-agents communication (#15556)

  Add `InterAgentCommunication` for v2 agent communication

  jif-oai · 2026-03-23 21:09:00 +00:00
  18f1a08bc9
• feat: structured multi-agent output (#15515)

  Send input now sends messages as assistant message and with this format:
  
  ```
  author: /root/worker_a
  recipient: /root/worker_a/tester
  other_recipients: []
  Content: bla bla bla. Actual content. Only text for now
  ```

  jif-oai · 2026-03-23 18:53:54 +00:00
  37ac0c093c
• chore(context) Include guardian approval context (#15366)

  ## Summary
  Include the guardian context in the developer message for approvals
  
  ## Testing
  - [x] Updated unit tests

  Dylan Hurd · 2026-03-21 16:31:22 +00:00
  0d9bb8ea58
• Split features into codex-features crate (#15253)

  - Split the feature system into a new `codex-features` crate.
  - Cut `codex-core` and workspace consumers over to the new config and
  warning APIs.
  
  Co-authored-by: Ahmed Ibrahim <219906144+aibrahim-oai@users.noreply.github.com>
  Co-authored-by: Codex <noreply@openai.com>

  Ahmed Ibrahim · 2026-03-19 20:12:07 -07:00
  2e22885e79
• [hooks] use a user message > developer message for prompt continuation (#14867)

  ## Summary
  
  Persist Stop-hook continuation prompts as `user` messages instead of
  hidden `developer` messages + some requested integration tests
  
  This is a followup to @pakrym 's comment in
  https://github.com/openai/codex/pull/14532 to make sure stop-block
  continuation prompts match training for turn loops
  
  - Stop continuation now writes `<hook_prompt hook_run_id="...">stop
  hook's user prompt<hook_prompt>`
  - Introduces quick-xml dependency, though we already indirectly depended
  on it anyway via syntect
  - This PR only has about 500 lines of actual logic changes, the rest is
  tests/schema
  
  ## Testing
  
  Example run (with a sessionstart hook and 3 stop hooks) - this shows
  context added by session start, then two stop hooks sending their own
  additional prompts in a new turn. The model responds with a single
  message addressing both. Then when that turn ends, the hooks detect that
  they just ran using `stop_hook_active` and decide not to infinite loop
  
  test files for this (unzip, move codex -> .codex):
  [codex.zip](https://github.com/user-attachments/files/26075806/codex.zip)
  
  ```
  › cats
  
  
  • Running SessionStart hook: lighting the observatory
  
  SessionStart hook (completed)
    warning: Hi, I'm a session start hook for wizard-tower (startup).
    hook context: A wimboltine stonpet is an exotic cuisine from hyperspace
  
  • Cats are tiny zen wizards, my friend: equal parts nap, mystery, and chaos. If you want, we can talk
    cat facts, cat breeds, cat names, or build something cat-themed in this repo.
  
  • Running Stop hook: checking the tower wards
  
  • Running Stop hook: sacking the guards
  
  • Running Stop hook: hiring the guards
  
  Stop hook (completed)
    warning: Wizard Tower Stop hook reviewed the completed reply (177 chars).
  
  Stop hook (blocked)
    warning: Wizard Tower Stop hook continuing conversation
    feedback: cook the stonpet
  
  Stop hook (blocked)
    warning: Wizard Tower Stop hook continuing conversation
    feedback: eat the cooked stonpet
  
  • Stonpet’s cooked, aloha style: flash-seared over a blue quasiflame, glazed with nebula salt, and
    rested until the hyperspace juices settle.
  
    Now we eat with gratitude, my friend. One mindful bite in, and the flavor is pure cosmic surf:
    smoky, bright, and totally out of this dimension.
  
  • Running Stop hook: checking the tower wards
  
  • Running Stop hook: sacking the guards
  
  • Running Stop hook: hiring the guards
  
  Stop hook (completed)
    warning: Wizard Tower Stop hook reviewed the completed reply (285 chars).
  
  Stop hook (completed)
    warning: Wizard Tower Stop hook saw a second pass and stayed calm to avoid a loop.
  
  Stop hook (completed)
    warning: Wizard Tower Stop hook saw a second pass and stayed calm to avoid a loop.
  ```

  Andrei Eternal · 2026-03-19 10:53:08 -07:00
  267499bed8
• Add notify to code-mode (#14842)

  Allows model to send an out-of-band notification.
  
  The notification is injected as another tool call output for the same
  call_id.

  pakrym-oai · 2026-03-18 09:37:13 -07:00
  606d85055f
• Apply argument comment lint across codex-rs (#14652)

  ## Why
  
  Once the repo-local lint exists, `codex-rs` needs to follow the
  checked-in convention and CI needs to keep it from drifting. This commit
  applies the fallback `/*param*/` style consistently across existing
  positional literal call sites without changing those APIs.
  
  The longer-term preference is still to avoid APIs that require comments
  by choosing clearer parameter types and call shapes. This PR is
  intentionally the mechanical follow-through for the places where the
  existing signatures stay in place.
  
  After rebasing onto newer `main`, the rollout also had to cover newly
  introduced `tui_app_server` call sites. That made it clear the first cut
  of the CI job was too expensive for the common path: it was spending
  almost as much time installing `cargo-dylint` and re-testing the lint
  crate as a representative test job spends running product tests. The CI
  update keeps the full workspace enforcement but trims that extra
  overhead from ordinary `codex-rs` PRs.
  
  ## What changed
  
  - keep a dedicated `argument_comment_lint` job in `rust-ci`
  - mechanically annotate remaining opaque positional literals across
  `codex-rs` with exact `/*param*/` comments, including the rebased
  `tui_app_server` call sites that now fall under the lint
  - keep the checked-in style aligned with the lint policy by using
  `/*param*/` and leaving string and char literals uncommented
  - cache `cargo-dylint`, `dylint-link`, and the relevant Cargo
  registry/git metadata in the lint job
  - split changed-path detection so the lint crate's own `cargo test` step
  runs only when `tools/argument-comment-lint/*` or `rust-ci.yml` changes
  - continue to run the repo wrapper over the `codex-rs` workspace, so
  product-code enforcement is unchanged
  
  Most of the code changes in this commit are intentionally mechanical
  comment rewrites or insertions driven by the lint itself.
  
  ## Verification
  
  - `./tools/argument-comment-lint/run.sh --workspace`
  - `cargo test -p codex-tui-app-server -p codex-tui`
  - parsed `.github/workflows/rust-ci.yml` locally with PyYAML
  
  ---
  
  * -> #14652
  * #14651

  Michael Bolin · 2026-03-16 16:48:15 -07:00
  b77fe8fefe
• Improve granular approval policy prompt (#14553)
  Jack Mousseau · 2026-03-13 10:42:17 -07:00
  59b588b8ec
• sending back imagaegencall response back to responseapi (#14558)

  Sending back the ResponseItem::ImageGenerationCall as is, because it is
  now supported from the API-side.

  Won Park · 2026-03-13 17:29:19 +00:00
  958f93f899
• Rename reject approval policy to granular (#14516)
  Jack Mousseau · 2026-03-12 16:38:04 -07:00
  b7dba72dbd
• feat: search_tool migrate to bring you own tool of Responses API (#14274)

  ## Why
  
  to support a new bring your own search tool in Responses
  API(https://developers.openai.com/api/docs/guides/tools-tool-search#client-executed-tool-search)
  we migrating our bm25 search tool to use official way to execute search
  on client and communicate additional tools to the model.
  
  ## What
  - replace the legacy `search_tool_bm25` flow with client-executed
  `tool_search`
  - add protocol, SSE, history, and normalization support for
  `tool_search_call` and `tool_search_output`
  - return namespaced Codex Apps search results and wire namespaced
  follow-up tool calls back into MCP dispatch

  Anton Panasenko · 2026-03-11 17:51:51 -07:00
  77b0c75267
• Add realtime start instructions config override (#14270)

  - add `realtime_start_instructions` config support
  - thread it into realtime context updates, schema, docs, and tests

  Ahmed Ibrahim · 2026-03-11 12:33:09 -07:00
  39c1bc1c68
• unifying all image saves to /tmp to bug-proof (#14149)

  image-gen feature will have the model saving to /tmp by default + at all
  times

  Won Park · 2026-03-11 12:33:08 -07:00
  722e8f08e1
• pass on save info to model + ui tweaks (#14123)

  Passing on more information to the model for context purposes, to
  streamline image-identification.

  Won Park · 2026-03-09 20:10:15 +00:00
  42f20a6845
• guardian initial feedback / tweaks (#13897)

  ## Summary
  - remove the remaining model-visible guardian-specific `on-request`
  prompt additions so enabling the feature does not change the main
  approval-policy instructions
  - neutralize user-facing guardian wording to talk about automatic
  approval review / approval requests rather than a second reviewer or
  only sandbox escalations
  - tighten guardian retry-context handling so agent-authored
  `justification` stays in the structured action JSON and is not also
  injected as raw retry context
  - simplify guardian review plumbing in core by deleting dead
  prompt-append paths and trimming some request/transcript setup code
  
  ## Notable Changes
  - delete the dead `permissions/approval_policy/guardian.md` append path
  and stop threading `guardian_approval_enabled` through model-facing
  developer-instruction builders
  - rename the experimental feature copy to `Automatic approval review`
  and update the `/experimental` snapshot text accordingly
  - make approval-review status strings generic across shell, patch,
  network, and MCP review types
  - forward real sandbox/network retry reasons for shell and unified-exec
  guardian review, but do not pass agent-authored justification as raw
  retry context
  - simplify `guardian.rs` by removing the one-field request wrapper,
  deduping reasoning-effort selection, and cleaning up transcript entry
  collection
  
  ## Testing
  - `just fmt`
  - full validation left to CI
  
  ---------
  
  Co-authored-by: Codex <noreply@openai.com>

  Charley Cunningham · 2026-03-09 09:25:24 -07:00
  f23fcd6ced
• Add guardian approval MVP (#13692)

  ## Summary
  - add the guardian reviewer flow for `on-request` approvals in command,
  patch, sandbox-retry, and managed-network approval paths
  - keep guardian behind `features.guardian_approval` instead of exposing
  a public `approval_policy = guardian` mode
  - route ordinary `OnRequest` approvals to the guardian subagent when the
  feature is enabled, without changing the public approval-mode surface
  
  ## Public model
  - public approval modes stay unchanged
  - guardian is enabled via `features.guardian_approval`
  - when that feature is on, `approval_policy = on-request` keeps the same
  approval boundaries but sends those approval requests to the guardian
  reviewer instead of the user
  - `/experimental` only persists the feature flag; it does not rewrite
  `approval_policy`
  - CLI and app-server no longer expose a separate `guardian` approval
  mode in this PR
  
  ## Guardian reviewer
  - the reviewer runs as a normal subagent and reuses the existing
  subagent/thread machinery
  - it is locked to a read-only sandbox and `approval_policy = never`
  - it does not inherit user/project exec-policy rules
  - it prefers `gpt-5.4` when the current provider exposes it, otherwise
  falls back to the parent turn's active model
  - it fail-closes on timeout, startup failure, malformed output, or any
  other review error
  - it currently auto-approves only when `risk_score < 80`
  
  ## Review context and policy
  - guardian mirrors `OnRequest` approval semantics rather than
  introducing a separate approval policy
  - explicit `require_escalated` requests follow the same approval surface
  as `OnRequest`; the difference is only who reviews them
  - managed-network allowlist misses that enter the approval flow are also
  reviewed by guardian
  - the review prompt includes bounded recent transcript history plus
  recent tool call/result evidence
  - transcript entries and planned-action strings are truncated with
  explicit `<guardian_truncated ... />` markers so large payloads stay
  bounded
  - apply-patch reviews include the full patch content (without
  duplicating the structured `changes` payload)
  - the guardian request layout is snapshot-tested using the same
  model-visible Responses request formatter used elsewhere in core
  
  ## Guardian network behavior
  - the guardian subagent inherits the parent session's managed-network
  allowlist when one exists, so it can use the same approved network
  surface while reviewing
  - exact session-scoped network approvals are copied into the guardian
  session with protocol/port scope preserved
  - those copied approvals are now seeded before the guardian's first turn
  is submitted, so inherited approvals are available during any immediate
  review-time checks
  
  ## Out of scope / follow-ups
  - the sandbox-permission validation split was pulled into a separate PR
  and is not part of this diff
  - a future follow-up can enable `serde_json` preserve-order in
  `codex-core` and then simplify the guardian action rendering further
  
  ---------
  
  Co-authored-by: Codex <noreply@openai.com>

  Charley Cunningham · 2026-03-07 05:40:10 -08:00
  e84ee33cc0
• image-gen-core (#13290)

  Core tool-calling for image-gen, handles requesting and receiving logic
  for images using response API

  Won Park · 2026-03-03 23:11:28 -08:00
  fa2306b303
• Add under-development original-resolution view_image support (#13050)

  ## Summary
  
  Add original-resolution support for `view_image` behind the
  under-development `view_image_original_resolution` feature flag.
  
  When the flag is enabled and the target model is `gpt-5.3-codex` or
  newer, `view_image` now preserves original PNG/JPEG/WebP bytes and sends
  `detail: "original"` to the Responses API instead of using the legacy
  resize/compress path.
  
  ## What changed
  
  - Added `view_image_original_resolution` as an under-development feature
  flag.
  - Added `ImageDetail` to the protocol models and support for serializing
  `detail: "original"` on tool-returned images.
  - Added `PromptImageMode::Original` to `codex-utils-image`.
    - Preserves original PNG/JPEG/WebP bytes.
    - Keeps legacy behavior for the resize path.
  - Updated `view_image` to:
  - use the shared `local_image_content_items_with_label_number(...)`
  helper in both code paths
    - select original-resolution mode only when:
      - the feature flag is enabled, and
      - the model slug parses as `gpt-5.3-codex` or newer
  - Kept local user image attachments on the existing resize path; this
  change is specific to `view_image`.
  - Updated history/image accounting so only `detail: "original"` images
  use the docs-based GPT-5 image cost calculation; legacy images still use
  the old fixed estimate.
  - Added JS REPL guidance, gated on the same feature flag, to prefer JPEG
  at 85% quality unless lossless is required, while still allowing other
  formats when explicitly requested.
  - Updated tests and helper code that construct
  `FunctionCallOutputContentItem::InputImage` to carry the new `detail`
  field.
  
  ## Behavior
  
  ### Feature off
  - `view_image` keeps the existing resize/re-encode behavior.
  - History estimation keeps the existing fixed-cost heuristic.
  
  ### Feature on + `gpt-5.3-codex+`
  - `view_image` sends original-resolution images with `detail:
  "original"`.
  - PNG/JPEG/WebP source bytes are preserved when possible.
  - History estimation uses the GPT-5 docs-based image-cost calculation
  for those `detail: "original"` images.
  
  
  #### [git stack](https://github.com/magus/git-stack-cli)
  - 👉 `1` https://github.com/openai/codex/pull/13050
  - ⏳ `2` https://github.com/openai/codex/pull/13331
  - ⏳ `3` https://github.com/openai/codex/pull/13049

  Curtis 'Fjord' Hawthorne · 2026-03-03 15:56:54 -08:00
  b92146d48b
• Record realtime close marker on replacement (#13058)

  ## Summary
  - record a realtime close developer message when a new realtime session
  replaces an active one
  - assert the replacement marker through the mocked responses request
  path
  
  ---------
  
  Co-authored-by: Codex <noreply@openai.com>
  Co-authored-by: Charles Cunningham <ccunningham@openai.com>

  Ahmed Ibrahim · 2026-03-01 13:54:12 -08:00
  0aeb55bf08
• Support multimodal custom tool outputs (#12948)

  ## Summary
  
  This changes `custom_tool_call_output` to use the same output payload
  shape as `function_call_output`, so freeform tools can return either
  plain text or structured content items.
  
  The main goal is to let `js_repl` return image content from nested
  `view_image` calls in its own `custom_tool_call_output`, instead of
  relying on a separate injected message.
  
  ## What changed
  
  - Changed `custom_tool_call_output.output` from `string` to
  `FunctionCallOutputPayload`
  - Updated freeform tool plumbing to preserve structured output bodies
  - Updated `js_repl` to aggregate nested tool content items and attach
  them to the outer `js_repl` result
  - Removed the old `js_repl` special case that injected `view_image`
  results as a separate pending user image message
  - Updated normalization/history/truncation paths to handle multimodal
  `custom_tool_call_output`
  - Regenerated app-server protocol schema artifacts
  
  ## Behavior
  
  Direct `view_image` calls still return a `function_call_output` with
  image content.
  
  When `view_image` is called inside `js_repl`, the outer `js_repl`
  `custom_tool_call_output` now carries:
  - an `input_text` item if the JS produced text output
  - one or more `input_image` items from nested tool results
  
  So the nested image result now stays inside the `js_repl` tool output
  instead of being injected as a separate message.
  
  ## Compatibility
  
  This is intended to be backward-compatible for resumed conversations.
  
  Older histories that stored `custom_tool_call_output.output` as a plain
  string still deserialize correctly, and older histories that used the
  previous injected-image-message flow also continue to resume.
  
  Added regression coverage for resuming a pre-change rollout containing:
  - string-valued `custom_tool_call_output`
  - legacy injected image message history
  
  
  #### [git stack](https://github.com/magus/git-stack-cli)
  - 👉 `1` https://github.com/openai/codex/pull/12948

  Curtis 'Fjord' Hawthorne · 2026-02-26 18:17:46 -08:00
  7e980d7db6
• core: bundle settings diff updates into one dev/user envelope (#12417)

  ## Summary
  - bundle contextual prompt injection into at most one developer message
  plus one contextual user message in both:
    - per-turn settings updates
    - initial context insertion
  - preserve `<model_switch>` across compaction by rebuilding it through
  canonical initial-context injection, instead of relying on
  strip/reattach hacks
  - centralize contextual user fragment detection in one shared definition
  table and reuse it for parsing/compaction logic
  - keep `AGENTS.md` in its natural serialized format:
    - `# AGENTS.md instructions for {dirname}`
    - `<INSTRUCTIONS>...</INSTRUCTIONS>`
  - simplify related tests/helpers and accept the expected snapshot/layout
  updates from bundled multi-part messages
  
  ## Why
  The goal is to converge toward a simpler, more intentional prompt shape
  where contextual updates are consistently represented as one developer
  envelope plus one contextual user envelope, while keeping parsing and
  compaction behavior aligned with that representation.
  
  ## Notable details
  - the temporary `SettingsUpdateEnvelope` wrapper was removed; these
  paths now return `Vec<ResponseItem>` directly
  - local/remote compaction no longer rely on model-switch strip/restore
  helpers
  - contextual user detection is now driven by shared fragment definitions
  instead of ad hoc matcher assembly
  - AGENTS/user instructions are still the same logical context; only the
  synthetic `<user_instructions>` wrapper was replaced by the natural
  AGENTS text format
  
  ## Testing
  - `just fmt`
  - `cargo test -p codex-app-server
  codex_message_processor::tests::extract_conversation_summary_prefers_plain_user_messages
  -- --exact`
  - `cargo test -p codex-core
  compact::tests::collect_user_messages_filters_session_prefix_entries
  --lib -- --exact`
  - `cargo test -p codex-core --test all
  'suite::compact::snapshot_request_shape_pre_turn_compaction_strips_incoming_model_switch'
  -- --exact`
  - `cargo test -p codex-core --test all
  'suite::compact_remote::snapshot_request_shape_remote_pre_turn_compaction_strips_incoming_model_switch'
  -- --exact`
  - `cargo test -p codex-core --test all
  'suite::client::includes_apps_guidance_as_developer_message_when_enabled'
  -- --exact`
  - `cargo test -p codex-core --test all
  'suite::client::includes_developer_instructions_message_in_request' --
  --exact`
  - `cargo test -p codex-core --test all
  'suite::client::includes_user_instructions_message_in_request' --
  --exact`
  - `cargo test -p codex-core --test all
  'suite::client::resume_includes_initial_messages_and_sends_prior_items'
  -- --exact`
  - `cargo test -p codex-core --test all
  'suite::review::review_input_isolated_from_parent_history' -- --exact`
  - `cargo test -p codex-exec --test all
  'suite::resume::exec_resume_last_respects_cwd_filter_and_all_flag' --
  --exact`
  - `cargo test -p core_test_support
  context_snapshot::tests::full_text_mode_preserves_unredacted_text --
  --exact`
  
  ## Notes
  - I also ran several targeted `compact`, `compact_remote`,
  `prompt_caching`, `model_visible_layout`, and `event_mapping` tests
  while iterating on prompt-shape changes.
  - I have not claimed a clean full-workspace `cargo test` from this
  environment because local sandbox/resource conditions have previously
  produced unrelated failures in large workspace runs.

  Charley Cunningham · 2026-02-26 00:12:08 -08:00
  07aefffb1f
• Raise image byte estimate for compaction token accounting (#12717)

  Increase `IMAGE_BYTES_ESTIMATE` from 340 bytes to 7,373 bytes so the
  existing 4-bytes/token heuristic yields an image estimate of ~1,844
  tokens instead of ~85. This makes auto-compaction more conservative for
  image-heavy transcripts and avoids underestimating context usage, which
  can otherwise cause compaction to fail when there is not enough free
  context remaining. The new value was chosen because that's the image
  resolution cap used for our latest models.
  
  Follow-up to [#12419](https://github.com/openai/codex/pull/12419).
  Refs [#11845](https://github.com/openai/codex/issues/11845).

  Eric Traut · 2026-02-24 16:11:38 -08:00
  8da40c9251
• feat(core) Introduce Feature::RequestPermissions (#11871)

  ## Summary
  Introduces the initial implementation of Feature::RequestPermissions.
  RequestPermissions allows the model to request that a command be run
  inside the sandbox, with additional permissions, like writing to a
  specific folder. Eventually this will include other rules as well, and
  the ability to persist these permissions, but this PR is already quite
  large - let's get the core flow working and go from there!
  
  <img width="1279" height="541" alt="Screenshot 2026-02-15 at 2 26 22 PM"
  src="https://github.com/user-attachments/assets/0ee3ec0f-02ec-4509-91a2-809ac80be368"
  />
  
  ## Testing
  - [x] Added tests
  - [x] Tested locally
  - [x] Feature

  Dylan Hurd · 2026-02-24 09:48:57 -08:00
  f6053fdfb3
• core: preserve constrained approval/sandbox policies in TurnContext (#12473)
  Michael Bolin · 2026-02-21 14:40:24 -08:00
  66d5d34e6e
• Improve token usage estimate for images (#12419)

  Fixes #11845.
  
  Adjust context/token estimation for inline image `data:*;base64,...`
  URLs so we
  do not count the raw base64 payload as model-visible text.
  
  What changed:
  - keep the existing JSON-length estimator as the baseline
  - detect only inline base64 `data:` image URLs in message and
  function-call
    output content items
  - subtract only the base64 payload bytes (preserving data URL prefix +
  JSON
    overhead)
  - add a fixed per-image estimate of 340 bytes (~85 tokens at the repo’s
    4-bytes/token heuristic)
  
  This avoids large overestimates from MCP image tool outputs while
  leaving normal
  image URLs (`https://`, `file://`, non-base64 `data:` URLs) unchanged.
  
  Tests:
  - message image data URL estimate regression
  - function-call output image data URL estimate regression
  - non-base64 image URLs unchanged
  - non-base64 `data:` URLs unchanged
  - `data:application/octet-stream;base64,...` adjusted
  - multiple inline images apply multiple fixed costs
  - text-only items unchanged

  Eric Traut · 2026-02-21 14:25:36 -08:00
  3586fcb802
• Fix compaction context reinjection and model baselines (#12252)

  ## Summary
  - move regular-turn context diff/full-context persistence into
  `run_turn` so pre-turn compaction runs before incoming context updates
  are recorded
  - after successful pre-turn compaction, rely on a cleared
  `reference_context_item` to trigger full context reinjection on the
  follow-up regular turn (manual `/compact` keeps replacement history
  summary-only and also clears the baseline)
  - preserve `<model_switch>` when full context is reinjected, and inject
  it *before* the rest of the full-context items
  - scope `reference_context_item` and `previous_model` to regular user
  turns only so standalone tasks (`/compact`, shell, review, undo) cannot
  suppress future reinjection or `<model_switch>` behavior
  - make context-diff persistence + `reference_context_item` updates
  explicit in the regular-turn path, with clearer docs/comments around the
  invariant
  - stop persisting local `/compact` `RolloutItem::TurnContext` snapshots
  (only regular turns persist `TurnContextItem` now)
  - simplify resume/fork previous-model/reference-baseline hydration by
  looking up the last surviving turn context from rollout lifecycle
  events, including rollback and compaction-crossing handling
  - remove the legacy fallback that guessed from bare `TurnContext`
  rollouts without lifecycle events
  - update compaction/remote-compaction/model-visible snapshots and
  compact test assertions (including remote compaction mock response
  shape)
  
  ## Why
  We were persisting incoming context items before spawning the regular
  turn task, which let pre-turn compaction requests accidentally include
  incoming context diffs without the new user message. Fixing that exposed
  follow-on baseline issues around `/compact`, resume/fork, and standalone
  tasks that could cause duplicate context injection or suppress
  `<model_switch>` instructions.
  
  This PR re-centers the invariants around regular turns:
  - regular turns persist model-visible context diffs/full reinjection and
  update the `reference_context_item`
  - standalone tasks do not advance those regular-turn baselines
  - compaction clears the baseline when replacement history may have
  stripped the referenced context diffs
  
  ## Follow-ups (TODOs left in code)
  - `TODO(ccunningham)`: fix rollback/backtracking baseline handling more
  comprehensively
  - `TODO(ccunningham)`: include pending incoming context items in
  pre-turn compaction threshold estimation
  - `TODO(ccunningham)`: inject updated personality spec alongside
  `<model_switch>` so some model-switch paths can avoid forced full
  reinjection
  - `TODO(ccunningham)`: review task turn lifecycle
  (`TurnStarted`/`TurnComplete`) behavior and emit task-start context
  diffs for task types that should have them (excluding `/compact`)
  
  ## Validation
  - `just fmt`
  - CI should cover the updated compaction/resume/model-visible snapshot
  expectations and rollout-hydration behavior
  - I did **not** rerun the full local test suite after the latest
  resume-lookup / rollout-persistence simplifications

  Charley Cunningham · 2026-02-20 23:13:08 -08:00
  bb0ac5be70
• Move previous turn context tracking into ContextManager history (#12179)

  ## Summary
  - add `previous_context_item: Option<TurnContextItem>` to
  `ContextManager`
  - expose session/state accessors for reading and updating the stored
  previous context item
  - switch settings diffing to use `TurnContextItem` instead of
  `TurnContext`
  - remove submission-loop local `previous_context` and persist the
  previous context item in history
  
  ## Testing
  - `just fmt`
  - `just fix -p codex-core`
  - `cargo test -p codex-core --test all model_switching::`
  - `cargo test -p codex-core --test all collaboration_instructions::`
  - `cargo test -p codex-core --test all personality::`
  - `cargo test -p codex-core --test all
  permissions_messages::permissions_message_not_added_when_no_change`

  Charley Cunningham · 2026-02-19 09:56:20 -08:00
  f2d5842ed1
• feat: sub-agent injection (#12152)

  This PR adds parent-thread sub-agent completion notifications and change
  the prompt of the model to prevent if from being confused

  jif-oai · 2026-02-19 11:32:10 +00:00
  2daa3fd44f
• Centralize context update diffing logic (#11807)

  ## Summary
  This PR centralizes model-visible state diffing for turn context updates
  into one module, while keeping existing behavior and call sites stable.
  
  ### What changed
  - Added `core/src/context_updates.rs` with the consolidated diffing
  logic for:
    - environment context updates
    - permissions/policy updates
    - collaboration mode updates
    - model-instruction switch updates
    - personality updates
  - Added `BuildSettingsUpdateItemsParams` so required dependencies are
  passed explicitly.
  - Updated `Session::build_settings_update_items` in `core/src/codex.rs`
  to delegate to the centralized module.
  - Reused the same centralized `personality_message_for` helper from
  initial-context assembly to avoid duplicated logic.
  - Registered the new module in `core/src/lib.rs`.
  
  ## Why
  This is a minimal, shippable step toward the model-visible-state design:
  all state diff decisions for turn-context update items now live in one
  place, improving reviewability and reducing drift risk without expanding
  scope.
  
  ## Behavior
  - Intended to be behavior-preserving.
  - No protocol/schema changes.
  - No call-site behavior changes beyond routing through the new
  centralized logic.
  
  ## Testing
  Ran targeted tests in this worktree:
  - `cargo test -p codex-core
  build_settings_update_items_emits_environment_item_for_network_changes`
  - `cargo test -p codex-core collaboration_instructions --test all`
  
  Both passed.
  
  ## Codex author
  `codex resume 019c540f-3951-7352-a3fa-6f07b834d4ce`

  Charley Cunningham · 2026-02-17 09:21:44 -08:00
  cab607befb
• Strip unsupported images from prompt history to guard against model switch (#11349)

  - Make `ContextManager::for_prompt` modality-aware and strip input_image
  content when the active model is text-only.
  - Added a test for multi-model -> text-only model switch

  Ahmed Ibrahim · 2026-02-10 11:58:00 -08:00
  5e01450963
• core: add focused diagnostics for remote compaction overflows (#11133)

  ## Summary
  - add targeted remote-compaction failure diagnostics in compact_remote
  logging
  - log the specific values needed to explain overflow timing:
    - last_api_response_total_tokens
    - estimated_tokens_of_items_added_since_last_successful_api_response
    - estimated_bytes_of_items_added_since_last_successful_api_response
    - failing_compaction_request_body_bytes
  - simplify breakdown naming and remove
  last_api_response_total_bytes_estimate (it was an approximation and not
  useful for debugging)
  
  ## Why
  When compaction fails with context_length_exceeded, we need concrete,
  low-ambiguity numbers that map directly to:
  1) what the API most recently reported, and
  2) what local history added since then.
  
  This keeps the failure logs actionable without adding broad, noisy
  metrics.
  
  ## Testing
  - just fmt
  - cargo test -p codex-core

  Charley Cunningham · 2026-02-09 12:42:20 -08:00
  9450cd9ce5
• core: account for all post-response items in auto-compact token checks (#11132)

  ## Summary
  - change compaction pre-check accounting to include all items added
  after the last model-generated item, not only trailing codex-generated
  outputs
  - use that boundary consistently in get_total_token_usage() and
  get_total_token_usage_breakdown()
  - update history tests to cover user/tool-output items after the last
  model item
  
  ## Why
  last_token_usage.total_tokens is API-reported for the last successful
  model response. After that point, local history may gain additional
  items (user messages, injected context, tool outputs). Compaction
  triggering must account for all of those items to avoid late compaction
  attempts that can overflow context.
  
  ## Testing
  - just fmt
  - cargo test -p codex-core

  Charley Cunningham · 2026-02-09 08:34:38 -08:00
  0883e5d3e5
• Fix remote compaction estimator/payload instruction small mismatch (#10692)

  ## Summary
  This PR fixes a deterministic mismatch in remote compaction where
  pre-trim estimation and the `/v1/responses/compact` payload could use
  different base instructions.
  
  Before this change:
  - pre-trim estimation used model-derived instructions
  (`model_info.get_model_instructions(...)`)
  - compact payload used session base instructions
  (`sess.get_base_instructions()`)
  
  After this change:
  - remote pre-trim estimation and compact payload both use the same
  `BaseInstructions` instance from session state.
  
  ## Changes
  - Added a shared estimator entry point in `ContextManager`:
  - `estimate_token_count_with_base_instructions(&self, base_instructions:
  &BaseInstructions) -> Option<i64>`
  - Kept `estimate_token_count(&TurnContext)` as a thin wrapper that
  resolves model/personality instructions and delegates to the new helper.
  - Updated remote compaction flow to fetch base instructions once and
  reuse it for both:
    - trim preflight estimation
    - compact request payload construction
  - Added regression coverage for parity and behavior:
    - unit test verifying explicit-base estimator behavior
  - integration test proving remote compaction uses session override
  instructions and trims accordingly
  
  ## Why this matters
  This removes a deterministic divergence source where pre-trim could
  think the request fits while the actual compact request exceeded context
  because its instructions were longer/different.
  
  ## Scope
  In scope:
  - estimator/payload base-instructions parity in remote compaction
  
  Out of scope:
  - retry-on-`context_length_exceeded`
  - compaction threshold/headroom policy changes
  - broader trimming policy changes
  
  ## Codex author:
  `codex fork 019c2b24-c2df-7b31-a482-fb8cf7a28559`

  Charley Cunningham · 2026-02-04 23:24:06 -08:00
  dc7007beaa