Commit Graph
100 Commits
Author SHA1 Message Date
pakrym-oai fd7f2aedc7 Handle response.incomplete (#11558)
Treat it same as error.
2026-02-12 00:11:38 -08:00
pakrym-oai d391f3e2f9 Hide the first websocket retry (#11548)
Sometimes connection needs to be quickly reestablished, don't produce an
error for that.
2026-02-11 22:48:13 -08:00
pakrym-oai b8e0d7594f Teach codex to test itself (#11531)
For fun and profit!
2026-02-11 20:03:19 -08:00
pakrym-oai 58eaa7ba8f Use slug in tui (#11519)
Display name is for VSCE and App, TUI uses lowercase everywhere.
2026-02-11 17:42:58 -08:00
pakrym-oai 703fb38d2a Make codex-sdk depend on openai/codex (#11503)
Do not bundle all binaries inside the SDK as it makes the package huge.
Instead depend on openai/codex
2026-02-11 16:20:10 -08:00
pakrym-oai d73de9c8ba Pump pings (#11413)
Keep processing ping even when the agent isn't actively running.

Otherwise the connection will drop.
2026-02-11 12:43:57 -08:00
pakrym-oai 2c3ce2048d Linkify feedback link (#11414)
Make it clickable
2026-02-11 11:21:03 -08:00
pakrym-oai eac5473114 Do not attempt to append after response.completed (#11402)
Completed responses are fully done, and new response must be created.
2026-02-11 07:45:17 -08:00
pakrym-oai 4473147985 Do not resend output items in incremental websockets connections (#11383)
In the incremental websocket output items are already part of the
context, no need to send them again and duplicate.
2026-02-10 19:38:08 -08:00
pakrym-oai c68999ee6d Prefer websocket transport when model opts in (#11386)
Summary
- add a `prefer_websockets` field to `ModelInfo`, defaulting to `false`
in all fixtures and constructors
- wire the new flag into websocket selection so models that opt in
always use websocket transport even when the feature gate is off

Testing
- Not run (not requested)
2026-02-10 18:50:48 -08:00
pakrym-oai bfd4e2112c Disable very flaky tests (#11394)
Collected from last 20 builds of main in
https://github.com/openai/codex/commits/main/.
2026-02-10 18:50:11 -08:00
pakrym-oai e4b5384539 Extract tool building (#11337)
Make it clear what input go into building tools and allow for easy reuse
for pre-warm request
2026-02-10 11:45:23 -08:00
pakrym-oai 0639c33892 Compare full request for websockets incrementality (#11343)
Tools can dynamically change mid-turn now. We need to be more thorough
about reusing incremental connections.
2026-02-10 19:14:36 +00:00
pakrym-oai 3322b99900 Remove ApiPrompt (#11265)
Keep things simple and build a full Responses API request request right
in the model client
2026-02-10 16:12:31 +00:00
pakrym-oai 086d02fb14 Try to stop small helper methods (#11203) 2026-02-09 20:01:30 +00:00
pakrym-oai 7044511ae8 Move warmup to the task level (#11216)
Instead of storing a special connection on the client level make the
regular task responsible for establishing a normal client session and
open a connection on it.

Then when the turn is started we pass in a pre-established session.
2026-02-09 11:58:53 -08:00
pakrym-oai ccd17374cb Move warmup to the task level (#11216)
Instead of storing a special connection on the client level make the
regular task responsible for establishing a normal client session and
open a connection on it.

Then when the turn is started we pass in a pre-established session.
2026-02-09 10:57:52 -08:00
pakrym-oai 6d08298f4e Fallback to HTTP on UPGRADE_REQUIRED (#10824)
Allow the server to trigger a connection downgrade in case the protocol
changes in incompatible ways.
2026-02-08 05:06:33 +00:00
pakrym-oai 8fe5066bcc Simplify pre-connect (#11040) 2026-02-07 15:52:03 -08:00
pakrym-oai dbe47ea01a Send beta header with websocket connects (#10727) 2026-02-05 15:05:02 -08:00
pakrym-oaiandJosh McKinney 0e8d359da9 Session-level model client (#10664)
Make ModelClient a session-scoped object.
Move state that is session level onto the client, and make state that is
per-turn explicit on corresponding methods.
Stop taking a huge Config object, instead only pass in values that are
actually needed.

---------

Co-authored-by: Josh McKinney <joshka@openai.com>
2026-02-04 16:58:48 -08:00
pakrym-oai 7f20357611 Stop client from being state carrier (#10595)
I'd like to make client session wide. This requires shedding all random
state it has to carry.
2026-02-04 09:05:37 -08:00
pakrym-oai 0efd33f7f4 Update tests to stop using sse_completed fixture (#10638)
Summary:
- replace the `sse_completed` fixture and related JSON template with
direct `responses::ev_completed` payload builders
- cascade the new SSE helpers through all affected core tests for
consistency and clarity
- remove legacy fixtures that were no longer needed once the helpers are
in place

Testing:
- Not run (not requested)
2026-02-04 08:38:06 -08:00
pakrym-oai 56ebfff1a8 Move metadata calculation out of client (#10589)
Model client shouldn't be responsible for this.
2026-02-03 21:59:13 -08:00
pakrym-oai 53d8474061 Ignore remote_compact_trims_function_call_history_to_fit_context_window on windows (#10474) 2026-02-02 22:47:38 -08:00
pakrym-oai bf87468c2b Restore status after preamble (#10465) 2026-02-02 20:35:50 -08:00
pakrym-oai cbfd2a37cc Trim compaction input (#10374)
Two fixes:

1. Include trailing tool output in the total context size calculation.
Otherwise when checking whether compaction should run we ignore newly
added outputs.
2. Trim trailing tool output/tool calls until we can fit the request
into the model context size. Otherwise the compaction endpoint will fail
to compact. We only trim items that can be reproduced again by the model
(tool calls, tool call outputs).
2026-02-02 19:03:11 -08:00
pakrym-oai 74327fa59c Select experimental features with space (#10281) 2026-02-02 11:35:11 -08:00
pakrym-oai 9d976962ec Add credits tooltip (#10274) 2026-02-02 10:06:43 -08:00
pakrym-oai 03fcd12e77 Do not append items on override turn context (#10354) 2026-02-01 18:51:26 -08:00
pakrym-oai 748141bdda Update announcement_tip.toml (#10267)
Extend the test for dev version
2026-01-30 14:14:29 -08:00
pakrym-oai 0fac2744f7 Hide /approvals from the slash-command list (#10265)
`/permissions` is the replacement. `/approvals` still available when
typing.
2026-01-30 22:12:50 +00:00
pakrym-oai 5f81e8e70b Fix main (#10262) 2026-01-30 21:54:05 +00:00
pakrym-oai aacd530a41 Update copy (#10256)
<img width="839" height="62" alt="image"
src="https://github.com/user-attachments/assets/ca987cdb-9e8c-403e-8856-a9b37baa7673"
/>
2026-01-30 12:57:19 -08:00
pakrym-oai fbb3a30953 Remove WebSocket wire format (#10179)
I'd like WireApi to go away (when chat is removed) and WebSockets is
still responses API just over a different transport.
2026-01-29 13:50:53 -08:00
pakrym-oai 3b1cddf001 Fall back to http when websockets fail (#10139)
I expect not all proxies work with websockets, fall back to http if
websockets fail.
2026-01-29 10:36:21 -08:00
pakrym-oai 998e88b12a Use test_codex more (#9961)
Reduces boilderplate.
2026-01-26 18:52:10 -08:00
pakrym-oaiandjif-oai b332482eb1 Mark collab as beta (#9834)
Co-authored-by: jif-oai <jif@openai.com>
2026-01-25 11:13:21 +01:00
pakrym-oai b511c38ddb Support end_turn flag (#9698)
Experimental flag that signals the end of the turn.
2026-01-22 17:27:48 +00:00
pakrym-oai 4d48d4e0c2 Revert "feat: support proxy for ws connection" (#9693)
Reverts openai/codex#9409
2026-01-22 15:57:18 +00:00
pakrym-oai 836f0343a3 Add tui.experimental_mode setting (#9656)
To simplify testing
2026-01-22 05:27:57 +00:00
pakrym-oai f2e1ad59bc Add websockets logging (#9633)
To help with debugging.
2026-01-21 21:35:38 +00:00
pakrym-oai 527b7b4c02 Feature to auto-enable websockets transport (#9578) 2026-01-20 20:32:06 -08:00
pakrym-oai e6d2ef432d Rename hierarchical_agents to child_agents_md (#9215)
Clearer name
2026-01-14 19:14:24 +00:00
pakrym-oai e958d0337e Log headers in trace mode (#9214)
To enable:

```
export RUST_LOG="warn,codex_=trace"
```

Sample: 
```
Request completed method=POST url=https://chatgpt.com/backend-api/codex/responses status=200 OK headers={"date": "Wed, 14 Jan 2026 18:21:21 GMT", "transfer-encoding": "chunked", "connection": "keep-alive", "x-codex-plan-type": "business", "x-codex-primary-used-percent": "3", "x-codex-secondary-used-percent": "6", "x-codex-primary-window-minutes": "300", "x-codex-primary-over-secondary-limit-percent": "0", "x-codex-secondary-window-minutes": "10080", "x-codex-primary-reset-after-seconds": "9944", "x-codex-secondary-reset-after-seconds": "171121", "x-codex-primary-reset-at": "1768424824", "x-codex-secondary-reset-at": "1768586001", "x-codex-credits-has-credits": "False", "x-codex-credits-balance": "", "x-codex-credits-unlimited": "False", "x-models-etag": "W/\"7a7ffbc83c159dbd7a2a73aaa9c91b7a\"", "x-oai-request-id": "ffedcd30-6d8a-4c4d-be10-8ebb23c142c8", "x-envoy-upstream-service-time": "417", "x-openai-proxy-wasm": "v0.1", "cf-cache-status": "DYNAMIC", "set-cookie": "__cf_bm=xFKeaMbWNbKO5ZX.K5cJBhj34OA1QvnF_3nkdMThjlA-1768414881-1.0.1.1-uLpsE_BDkUfcmOMaeKVQmv_6_2ytnh_R3lO_il5N5K3YPQEkBo0cOMTdma6bK0Gz.hQYcIesFwKIJht1kZ9JKqAYYnjgB96hF4.sii2U3cE; path=/; expires=Wed, 14-Jan-26 18:51:21 GMT; domain=.chatgpt.com; HttpOnly; Secure; SameSite=None", "report-to": "{\"endpoints\":[{\"url\":\"https:\/\/a.nel.cloudflare.com\/report\/v4?s=4Kc7g4zUhKkIm3xHuB6ba4jyIUqqZ07ETwIPAYQASikRjA8JesbtUKDP9tSrZ5PnzWldaiSz5dZVQFI579LEsCMlMUSelTvmyQ8j4FbFDawi%2FprWZ5iRePiaSalr\"}],\"group\":\"cf-nel\",\"max_age\":604800}", "nel": "{\"success_fraction\":0.01,\"report_to\":\"cf-nel\",\"max_age\":604800}", "strict-transport-security": "max-age=31536000; includeSubDomains; preload", "x-content-type-options": "nosniff", "cross-origin-opener-policy": "same-origin-allow-popups", "referrer-policy": "strict-origin-when-cross-origin", "server": "cloudflare", "cf-ray": "9bdf270adc7aba3a-SEA"} version=HTTP/1.1
```
2026-01-14 18:38:12 +00:00
pakrym-oai 92472e7baa Use current model for review (#9179)
Instead of having a hard-coded default review model, use the current
model for running `/review` unless one is specified in the config.

Also inherit current reasoning effort
2026-01-14 08:59:41 -08:00
pakrym-oai 5a70b1568f WebSocket test server script (#9175)
Very simple test server to test Responses API WebSocket transport
integration.
2026-01-13 16:21:14 -08:00
pakrym-oai 9f8d3c14ce Fix flakiness in WebSocket tests (#9169)
The connection was being added to the list after the WebSocket response
was sent.

So the test can sometimes race and observe connections before the list
was updated.

After this change, connection and request is added to the list before
the response is sent.
2026-01-13 15:09:59 -08:00
pakrym-oai 2d56519ecd Support response.done and add integration tests (#9129)
The agent loop using a persistent incremental web socket connection.
2026-01-13 16:12:30 +00:00
pakrym-oai e726a82c8a Websocket append support (#9128)
Support an incremental append request in websocket transport.
2026-01-13 06:07:13 +00:00
pakrym-oai d75626ad99 Reuse websocket connection (#9127)
Reuses the connection but still sends full requests.
2026-01-13 03:30:09 +00:00
pakrym-oai 490c1c1fdd Add model client sessions (#9102)
Maintain a long-running session.
2026-01-13 01:15:56 +00:00
pakrym-oai 3a6a43ff5c Extract single responses SSE event parsing (#9114)
To be reused in WebSockets parsing.
2026-01-12 13:59:51 -08:00
pakrym-oai 5dfa780f3d Remove unused conversation_id header (#9107)
It's an exact copy of session_id
2026-01-12 21:01:07 +00:00
pakrym-oai 1a0e2e612b Delete announcement_tip.toml (#9003) 2026-01-09 14:47:46 -08:00
pakrym-oai acfd94f625 Add hierarchical agent prompt (#8996) 2026-01-09 13:47:37 -08:00
pakrym-oai cabf85aa18 Log unhandled sse events (#8949) 2026-01-09 12:36:07 -08:00
pakrym-oai 62a73b6d58 Attempt to reload auth as a step in 401 recovery (#8880)
When authentication fails, first attempt to reload the auth from file
and then attempt to refresh it.
2026-01-08 15:06:44 -08:00
pakrym-oai 634764ece9 Immutable CodexAuth (#8857)
Historically we started with a CodexAuth that knew how to refresh it's
own tokens and then added AuthManager that did a different kind of
refresh (re-reading from disk).

I don't think it makes sense for both `CodexAuth` and `AuthManager` to
be mutable and contain behaviors.

Move all refresh logic into `AuthManager` and keep `CodexAuth` as a data
object.
2026-01-08 11:43:56 -08:00
pakrym-oai 018de994b0 Stop using AuthManager as the source of codex_home (#8846) 2026-01-07 18:56:20 +00:00
pakrym-oai fedcb8f63c Move tests below auth manager (#8840)
To simplify future diffs
2026-01-07 17:36:44 +00:00
pakrym-oai 1b5095b5d1 Attach more tags to feedback submissions (#8688)
Attach more tags to sentry feedback so it's easier to classify and debug
without having to scan through logs.

Formatting isn't amazing but it's a start.
<img width="1234" height="276" alt="image"
src="https://github.com/user-attachments/assets/521a349d-f627-4051-b511-9811cd5cd933"
/>
2026-01-02 16:51:03 -08:00
pakrym-oai 7078a0b676 Log compaction request bodies (#8676)
We already log request bodies for normal requests, logging for
compaction helps with debugging.
2026-01-02 11:27:37 -08:00
pakrym-oai 96fdbdd434 Add ExecPolicyManager (#8349)
Move exec policy management into services to keep turn context
immutable.
2025-12-22 09:59:32 -08:00
pakrym-oai 8f0b383621 model list (#8286)
<img width="200" alt="7ff2254b-e96f-42fc-8232-b4e76cb26248"
src="https://github.com/user-attachments/assets/1f56799d-e2cd-4b69-9290-854943f7c6b6"
/>
2025-12-18 14:13:49 -08:00
pakrym-oai b3ddd50eee Remote compact for API-key users (#7835) 2025-12-12 10:05:02 -08:00
pakrym-oai bb8fdb20dc Revert "Only show Worked for after the final assistant message" (#7884)
Reverts openai/codex#7854
2025-12-11 09:11:42 -08:00
pakrym-oai 83aac0f985 Only show Worked for after the final assistant message (#7854)
Before:
<img width="1908" height="246" alt="image"
src="https://github.com/user-attachments/assets/f4d5993a-8d37-4982-a6fd-d37f449215b2"
/>
After:
<img width="1102" height="586" alt="image"
src="https://github.com/user-attachments/assets/e833140d-690a-4c33-8bc7-e2b69b9dc92d"
/>
2025-12-10 21:13:13 -08:00
pakrym-oai eb2e5458cc Disable ansi codes in tui log file (#7836) 2025-12-10 13:56:48 -08:00
pakrym-oai 4b684c53ae Remove conversation_id and bring back request ID logging (#7830) 2025-12-10 10:44:12 -08:00
pakrym-oai 967d063f4b parse rg | head a search (#7797) 2025-12-09 18:30:16 -08:00
pakrym-oaiandPavel 164265bed1 Vendor ConPtySystem (#7656)
The repo we were depending on is very large and we need very small part
of it.

---------

Co-authored-by: Pavel <pavel@krymets.com>
2025-12-09 17:23:51 +00:00
pakrym-oai ac5fa6baf8 Do not emit start/end events for write stdin (#7561) 2025-12-08 15:23:02 -08:00
pakrym-oai f1b7cdc3bd Use shared check sandboxing (#7547) 2025-12-04 08:34:09 -08:00
pakrym-oai c4e18f1b63 Slightly better status display for unified exec (#7563)
Trim bash -lc
2025-12-04 08:32:54 -08:00
pakrym-oai 8b1e397211 Add request logging back (#7471)
Having full requests helps debugging
2025-12-02 07:57:55 -08:00
pakrym-oai e52cc38dfd Use use_model (#7121) 2025-11-21 22:10:52 +00:00
pakrym-oai ab5972d447 Support all types of search actions (#7061)
Fixes the 

```
{
  "error": {
    "message": "Invalid value: 'other'. Supported values are: 'search', 'open_page', and 'find_in_page'.",
    "type": "invalid_request_error",
    "param": "input[150].action.type",
    "code": "invalid_value"
  }
```
error.


The actual-actual fix here is supporting absent `query` parameter.
2025-11-20 20:45:28 -08:00
pakrym-oai 767b66f407 Migrate coverage to shell_command (#7042) 2025-11-21 03:44:00 +00:00
pakrym-oai 830ab4ce20 Support full powershell paths in is_safe_command (#7055)
New shell implementation always uses full paths.
2025-11-20 19:29:15 -08:00
pakrym-oai 856f97f449 Delete shell_command feature (#7024) 2025-11-20 14:14:56 -08:00
pakrym-oai 52d0ec4cd8 Delete tiktoken-rs (#7018) 2025-11-20 11:15:04 -08:00
pakrym-oai 30ca89424c Always fallback to real shell (#6953)
Either cmd.exe or `/bin/sh`.
2025-11-20 10:58:46 -08:00
pakrym-oai 92e3046733 Single pass truncation (#6914) 2025-11-19 16:56:37 -08:00
pakrym-oai 64ae9aa3c3 Keep gpt-5.1-codex the default (#6922) 2025-11-19 11:08:10 -08:00
pakrym-oai 75f38f16dd Run remote auto compaction (#6879) 2025-11-19 00:43:58 -08:00
pakrym-oai ee0484a98c shell_command returns freeform output (#6860)
Instead of returning structured out and then re-formatting it into
freeform, return the freeform output from shell_command tool.

Keep `shell` as the default tool for GPT-5.
2025-11-18 23:38:43 -08:00
pakrym-oai cac0a6a29d Remote compaction on by-default (#6866) 2025-11-19 02:21:57 +00:00
pakrym-oai 018a2d2e50 Ignore unified_exec_respects_workdir_override (#6693) 2025-11-14 15:00:31 -08:00
pakrym-oai cfcc87a953 Order outputs before inputs (#6691)
For better caching performance all output items should be rendered in
the order they were produced before all new input items (for example,
all function_call before all function_call_output).
2025-11-14 14:54:11 -08:00
pakrym-oai 4ba562d2dd Add test timeout (#6612)
Add an overall test timeout of 30s.
2025-11-14 09:30:37 -08:00
pakrym-oai 6c384eb9c6 tests: replace mount_sse_once_match with mount_sse_once for SSE mocking (#6640) 2025-11-13 18:04:05 -08:00
pakrym-oai 7b027e7536 Revert "Revert "Overhaul shell detection and centralize command generation for unified exec"" (#6607)
Reverts openai/codex#6606
2025-11-13 16:45:17 -08:00
pakrym-oai 547be54ee8 Only list failed tests (#6619)
Makes output easier to parse
2025-11-13 13:50:33 -08:00
pakrym-oai c95bd345ea Enable close-stale-contributor-prs.yml workflow (#6615)
Tested on https://github.com/openai/codex/pull/3036
2025-11-13 11:50:54 -08:00
pakrym-oai 0792a7953d Update default yield time (#6610)
10s for exec and 250ms for write_stdin
2025-11-13 10:24:41 -08:00
pakrym-oai 6cda3de3a4 Close stale PRs workflow (#6594)
Closes stale PRs from OpenAI employees.
2025-11-13 10:01:51 -08:00
pakrym-oai 041d6ad902 Migrate prompt caching tests to test_codex (#6605)
To hopefully fix the flakiness
2025-11-13 09:19:38 -08:00
pakrym-oai e6995174c1 Revert "Overhaul shell detection and centralize command generation for unified exec" (#6606)
Reverts openai/codex#6577
2025-11-13 08:43:00 -08:00
pakrym-oaiandPavel Krymets d28e912214 Overhaul shell detection and centralize command generation for unified exec (#6577)
This fixes command display for unified exec. All `cd`s and `ls`es are
now parsed.

<img width="452" height="237" alt="image"
src="https://github.com/user-attachments/assets/ce92d81f-f74c-485a-9b34-1eaa29290ec6"
/>

Deletes a ton of tests that were doing nothing from shell.rs.

---------

Co-authored-by: Pavel Krymets <pavel@krymets.com>
2025-11-13 08:28:09 -08:00