fix(server): Responses API emits incomplete status on truncation (#19)#82
Merged
Merged
Conversation
When generation hits `STOP_TYPE_LIMIT` (max_output_tokens / ctx-size cap), the OAI Responses code paths hardcoded `"status": "completed"` everywhere — top-level response, per-message output items, function_call items, and the streaming `response.completed` SSE event. Agentic clients (Codex CLI, etc.) couldn't tell a finished response from a truncated one and ended up feeding partial output back into conversation history, triggering infinite retry loops on JSON-parse failures (issue #19, Phase 2). Per the OAI Responses spec, branch on the stop type in: * `server_task_result_cmpl_final::to_json_oaicompat_resp()` — emit `"status": "incomplete"` on the top-level response, all output items inherit the same status, plus `"incomplete_details": {"reason": "max_output_tokens"}` at the top level. * `to_json_oaicompat_resp_stream()` — same mapping on the per-item statuses, plus the final SSE event becomes `response.incomplete` (vs `response.completed`) with `incomplete_details` on the payload. Doesn't address Phase 2 of the issue (HTTP 400 + actionable message from `func_args_not_string`) — that requires typed exception plumbing through common/chat.cpp into the server error path. Phase 1 alone prevents the cascade in the first place: clients see truncation as truncation, not as a malformed completed response. Test coverage in test_compat_oai_responses.py: * `test_responses_truncation_emits_incomplete_status` — non-streaming: `max_output_tokens: 2` on tinyllama2 reliably trips STOP_TYPE_LIMIT; assert status=incomplete + incomplete_details + per-item status. * `test_responses_truncation_stream_emits_incomplete_event` — streaming: same setup, verify a `response.incomplete` event arrives with the same payload shape.
This was referenced Jun 5, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Closes Phase 1 of #19: when generation hits
STOP_TYPE_LIMIT(max_output_tokens / ctx-size cap), the OAI Responses code paths used to hardcode"status": "completed"on the top-level response, all output items, and the streamingresponse.completedSSE event. Agentic clients (Codex CLI, etc.) couldn't tell a finished response from a truncated one — they fed partial output back into conversation history, triggering JSON-parse-failure 500s on the next request and infinite retry loops.Per the OAI Responses spec:
statusflips to"incomplete"on truncation.incomplete_details: { reason: "max_output_tokens" }field.status(message / reasoning / function_call) inherits the same value, so clients can detect partial tool_calls / partial messages at the per-item level.response.incomplete(instead ofresponse.completed), with the same payload shape.What this does not do
Phase 2 of #19 (HTTP 400 + actionable message from
func_args_not_string) is intentionally out of scope. That requires typed-exception plumbing throughcommon/chat.cppinto the server error path — a separate, bigger change. Phase 1 alone prevents the cascade in the first place: once clients see truncation as truncation, they don't retry with malformed history.Test plan
tools/server/tests/unit/test_compat_oai_responses.py::test_responses_truncation_emits_incomplete_status— non-streaming repro withmax_output_tokens: 2on tinyllama2 (reliably trips STOP_TYPE_LIMIT). Asserts top-levelstatus=incomplete+incomplete_details.reason=max_output_tokens+ per-item status.test_responses_truncation_stream_emits_incomplete_event— streaming repro. Verifies aresponse.incompleteevent arrives with the same payload shape.test_responses_with_openai_library/test_responses_stream_with_openai_librarytests still pass (no happy-path regression).Files touched
tools/server/server-task.cpp— bothto_json_oaicompat_respandto_json_oaicompat_resp_streamtools/server/tests/unit/test_compat_oai_responses.py— 2 new test cases🤖 Generated with Claude Code