mirror of
https://github.com/openai/codex.git
synced 2026-11-09 20:50:33 +00:00
Dylan Hurd
8703d32bd0
## What changed - Pass goal objectives as untrusted `goal_objective` context instead of embedding them in continuation instructions. Direct the model to use `get_goal` if the objective is missing or truncated. - Inject newly created objectives into their originating active turn without dropping unrelated client context. Render them as tool outputs when `UntrustedAsToolOutput` is enabled. - Queue additional context together with automatic turn input so both reach the first model request. ## Testing Extend app-server coverage to verify that model-written objectives appear as tool outputs at creation and continuation, without duplicating them in user messages. Update steering and Python SDK tests to check objective preservation and delivery alongside continuation instructions. GitOrigin-RevId: 6f44f6d6c895bb4dc378c9cf84d6470ccc885c67
96 lines
3.6 KiB
Python
96 lines
3.6 KiB
Python
from app_server_harness import (
|
|
AppServerHarness,
|
|
ev_assistant_message,
|
|
ev_completed,
|
|
ev_function_call,
|
|
ev_response_created,
|
|
sse,
|
|
)
|
|
from app_server_helpers import agent_message_texts
|
|
|
|
from openai_codex import Codex
|
|
from openai_codex._goal import _GoalNotificationStream
|
|
from openai_codex._run import _collect_turn_result
|
|
from openai_codex.generated.notification_registry import notification_turn_id
|
|
from openai_codex.generated.v2_all import TurnStatus
|
|
|
|
|
|
def test_private_goal_operation_coalesces_runtime_continuations(tmp_path) -> None:
|
|
"""The private engine should expose automatic continuations as one turn."""
|
|
with AppServerHarness(tmp_path) as harness:
|
|
harness.responses.enqueue_assistant_message(
|
|
"Initial pass complete.",
|
|
response_id="goal-initial",
|
|
)
|
|
harness.responses.enqueue_sse(
|
|
sse(
|
|
[
|
|
ev_response_created("goal-complete-tool"),
|
|
ev_function_call(
|
|
"call-goal-complete",
|
|
"update_goal",
|
|
'{"status":"complete"}',
|
|
),
|
|
ev_completed("goal-complete-tool"),
|
|
]
|
|
)
|
|
)
|
|
harness.responses.enqueue_sse(
|
|
sse(
|
|
[
|
|
ev_response_created("goal-final"),
|
|
ev_assistant_message("msg-goal-final", "Goal complete."),
|
|
ev_completed("goal-final"),
|
|
]
|
|
)
|
|
)
|
|
|
|
with Codex(config=harness.app_server_config()) as codex:
|
|
thread = codex.thread_start()
|
|
state, turn_id = codex._client.start_goal_operation( # noqa: SLF001
|
|
thread.id,
|
|
"Improve benchmark coverage",
|
|
)
|
|
stream = _GoalNotificationStream(
|
|
state,
|
|
state.next_notification,
|
|
lambda: codex._client.unregister_goal_operation(state), # noqa: SLF001
|
|
lambda: codex._client.cancel_goal_operation(state), # noqa: SLF001
|
|
)
|
|
events = list(stream)
|
|
result = _collect_turn_result(iter(events), turn_id=turn_id)
|
|
routes = codex._client._router._goal_operations.copy() # noqa: SLF001
|
|
requests = harness.responses.wait_for_requests(3)
|
|
|
|
lifecycle = [event.method for event in events if event.method.startswith("turn/")]
|
|
routed_ids = [
|
|
routed_id
|
|
for event in events
|
|
if (routed_id := notification_turn_id(event.payload)) is not None
|
|
]
|
|
assert {
|
|
"lifecycle": lifecycle,
|
|
"routed_ids": routed_ids,
|
|
"result": (result.id, result.status, result.final_response),
|
|
"messages": agent_message_texts(events),
|
|
"request_count": len(requests),
|
|
"objective_reached_model": (
|
|
"<external_goal_objective>Improve benchmark coverage</external_goal_objective>"
|
|
in requests[0].message_input_texts("user")
|
|
),
|
|
"steering_reached_model": any(
|
|
"Continue working toward the active thread goal." in text
|
|
for text in requests[0].message_input_texts("user")
|
|
),
|
|
"routes_after_completion": routes,
|
|
} == {
|
|
"lifecycle": ["turn/started", "turn/completed"],
|
|
"routed_ids": [turn_id] * len(routed_ids),
|
|
"result": (turn_id, TurnStatus.completed, "Goal complete."),
|
|
"messages": ["Initial pass complete.", "Goal complete."],
|
|
"request_count": 3,
|
|
"objective_reached_model": True,
|
|
"steering_reached_model": True,
|
|
"routes_after_completion": {},
|
|
}
|