Commit graph

14 commits

Author SHA1 Message Date
Ahmed Ibrahim
b11e96fb04
Act on reasoning-included per turn (#9402)
- Reset reasoning-included flag each turn and update compaction test
2026-01-19 11:23:25 -08:00
Fouad Matin
93a5e0fe1c
fix(codex-api): treat invalid_prompt as non-retryable (#9400)
**Goal**: Prevent response.failed events with `invalid_prompt` from
being treated as retryable errors so the UI shows the actual error
message instead of continually retrying.

**Before**: Codex would continue to retry despite the prompt being
marked as disallowed
**After**: Codex will stop retrying once prompt is marked disallowed
2026-01-16 22:22:08 -08:00
Ahmed Ibrahim
ebdd8795e9
Turn-state sticky routing per turn (#9332)
- capture the header from SSE/WS handshakes, store it per
ModelClientSession using `Oncelock`, echo it on turn-scoped requests,
and add SSE+WS integration tests for within-turn persistence +
cross-turn reset.

- keep `x-codex-turn-state` sticky within a user turn to maintain
routing continuity for retries/tool follow-ups.
2026-01-16 09:30:11 -08:00
Michael Bolin
3728db11b8
fix: eliminate unnecessary clone() for each SSE event (#9238)
Given how many SSE events we get, seems worth fixing.
2026-01-15 00:06:09 +00:00
Celia Chen
02f67bace8
fix: Emit response.completed immediately for Responses SSE (#9170)
we see windows test failures like this:
https://github.com/openai/codex/actions/runs/20930055601/job/60138344260.

The issue is that SSE connections sometimes remain open after the
completion event esp. for windows. We should emit the completion event
and return immediately. this is consistent with the protocol:

> The Model streams responses back in an SSE, which are collected until
"completed" message and the SSE terminates

from
https://github.com/openai/codex/blob/dev/cc/fix-windows-test/codex-rs/docs/protocol_v1.md#L37.

this helps us achieve parity with responses websocket logic here:
https://github.com/openai/codex/blob/dev/cc/fix-windows-test/codex-rs/codex-api/src/endpoint/responses_websocket.rs#L220-L227.
2026-01-14 10:05:00 -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
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
cabf85aa18
Log unhandled sse events (#8949) 2026-01-09 12:36:07 -08:00
Josh McKinney
bba5e5e0d4
fix(codex-api): handle Chat Completions DONE sentinel (#8708)
Context
- This code parses Server-Sent Events (SSE) from the legacy Chat
Completions streaming API (wire_api = "chat").
- The upstream protocol terminates a stream with a final sentinel event:
data: [DONE].
- Some of our test stubs/helpers historically end the stream with data:
DONE (no brackets).

How this was found
- GitHub Actions on Windows failed in codex-app-server integration tests
with wiremock verification errors (expected multiple POSTs, got 1).

Diagnosis
- The job logs included: codex_api::sse::chat: Failed to parse
ChatCompletions SSE event ... data: DONE.
- eventsource_stream surfaces the sentinel as a normal SSE event; it
does not automatically close the stream.
- The parser previously attempted to JSON-decode every data: payload.
The sentinel is not JSON, so we logged and skipped it, then continued
polling.
- On servers that keep the HTTP connection open after emitting the
sentinel (notably wiremock on Windows), skipping the sentinel meant we
never emitted ResponseEvent::Completed.
- Higher layers wait for completion before progressing (emitting
approval requests and issuing follow-up model calls), so the test never
reached the subsequent requests and wiremock panicked when its
expected-call count was not met.

Fix
- Treat both data: [DONE] and data: DONE as explicit end-of-stream
sentinels.
- When a sentinel is seen, flush any pending assistant/reasoning items
and emit ResponseEvent::Completed once.

Tests
- Add a regression unit test asserting we complete on the sentinel even
if the underlying connection is not closed.
2026-01-05 09:29:42 -08:00
Ahmed Ibrahim
66b7c673e9
Refresh on models etag mismatch (#8491)
- Send models etag
- Refresh models on 412
- This wires `ModelsManager` to `ModelFamily` so we don't mutate it
mid-turn
2026-01-01 11:41:16 -08:00
jif-oai
5f80ad6da8
fix: chat completion with parallel tool call (#7634) 2025-12-05 10:20:36 -08:00
jif-oai
6736d1828d
fix: sse for chat (#7594) 2025-12-04 16:46:56 -08:00
jif-oai
4502b1b263
chore: proper client extraction (#6996) 2025-11-25 18:06:12 +00:00