> This page is for version 0.4.0. > For other versions, use one of these documentation indexes: > - Main (default): https://docs.nvidia.com/nemo/gym/main/llms.txt > - 0.7.0: https://docs.nvidia.com/nemo/gym/v0.7.0/llms.txt > - 0.6.0: https://docs.nvidia.com/nemo/gym/v0.6.0/llms.txt > - 0.5.1: https://docs.nvidia.com/nemo/gym/v0.5.1/llms.txt > - 0.5.0: https://docs.nvidia.com/nemo/gym/v0.5.0/llms.txt > - 0.4.0: https://docs.nvidia.com/nemo/gym/v0.4.0/llms.txt > - 0.3.0: https://docs.nvidia.com/nemo/gym/v0.3.0/llms.txt > - 0.2.1: https://docs.nvidia.com/nemo/gym/v0.2.1/llms.txt > For clean Markdown of any page, append .md to the page URL. > For a complete documentation index, see https://docs.nvidia.com/nemo/gym/llms.txt. > For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.nvidia.com/nemo/gym/_mcp/server. # nemo_gym.trajectory.trajectory_record Shared contracts for rollout observations and trajectories. ## Module Contents ### Classes | Name | Description | | ----------------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------- | | [`AgentEpisode`](#nemo_gym-trajectory-trajectory_record-AgentEpisode) | An Agent response and the observations available at its execution boundary. | | [`AgentInvocation`](#nemo_gym-trajectory-trajectory_record-AgentInvocation) | One agent activation or subagent invocation observed by a harness. | | [`AgentObservationBundle`](#nemo_gym-trajectory-trajectory_record-AgentObservationBundle) | Normalized observations returned by one Agent Server for one rollout. | | [`ContextCompactionObservation`](#nemo_gym-trajectory-trajectory_record-ContextCompactionObservation) | An explicit context-compaction event reported by the Agent harness. | | [`GradingOperationObservation`](#nemo_gym-trajectory-trajectory_record-GradingOperationObservation) | A source-reviewed grading calculation actually executed inside a verifier invocation. | | [`ModelCallRef`](#nemo_gym-trajectory-trajectory_record-ModelCallRef) | Stable identifiers an Agent integration can observe for one model call. | | [`ObservationGap`](#nemo_gym-trajectory-trajectory_record-ObservationGap) | A fact that the selected integration could not observe or join exactly. | | [`ObservationModel`](#nemo_gym-trajectory-trajectory_record-ObservationModel) | - | | [`OutputBudgetRefit`](#nemo_gym-trajectory-trajectory_record-OutputBudgetRefit) | A model call the server sent again with its output budget lowered, because the budget it requested exceeded | | [`SandboxObservation`](#nemo_gym-trajectory-trajectory_record-SandboxObservation) | Outcome and lifetime resource usage reported by a sandbox-owning harness. | | [`ToolCallObservation`](#nemo_gym-trajectory-trajectory_record-ToolCallObservation) | Raw tool evidence; span and parent IDs are local to its producer until imported. | | [`TrajectoryModelCall`](#nemo_gym-trajectory-trajectory_record-TrajectoryModelCall) | - | | [`TrajectoryRecord`](#nemo_gym-trajectory-trajectory_record-TrajectoryRecord) | - | | [`TrajectoryResponseMetadata`](#nemo_gym-trajectory-trajectory_record-TrajectoryResponseMetadata) | - | | [`TrajectoryTokenStats`](#nemo_gym-trajectory-trajectory_record-TrajectoryTokenStats) | - | | [`TrajectoryToolCall`](#nemo_gym-trajectory-trajectory_record-TrajectoryToolCall) | Canonical tool attempt with qualified span IDs and model-visible output. | | [`TrajectoryTurn`](#nemo_gym-trajectory-trajectory_record-TrajectoryTurn) | - | | [`VerifierGradingRecord`](#nemo_gym-trajectory-trajectory_record-VerifierGradingRecord) | The grade a verifier returned without asking any model, recorded by the shared verify path. | | [`VerifierInvocation`](#nemo_gym-trajectory-trajectory_record-VerifierInvocation) | One verify() execution, its outcome, and all model attempts made while grading. | ### Functions | Name | Description | | ----------------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------- | | [`exclude_verifier_model_calls`](#nemo_gym-trajectory-trajectory_record-exclude_verifier_model_calls) | Exclude explicitly verifier-owned attempts from agent and step joins. | | [`failure_observations`](#nemo_gym-trajectory-trajectory_record-failure_observations) | Keep the existing typed observation slots across an unsuccessful HTTP reply. | | [`grading_result_binding`](#nemo_gym-trajectory-trajectory_record-grading_result_binding) | Bind the actual graded response and numeric reward/components, without exporting their payload. | | [`join_agent_model_calls`](#nemo_gym-trajectory-trajectory_record-join_agent_model_calls) | Join captured calls to agent invocations and their compactions without guessing ownership. | | [`join_assistant_message_calls`](#nemo_gym-trajectory-trajectory_record-join_assistant_message_calls) | Join persisted messages to HTTP attempts using session, message and model-server IDs. | | [`tool_parent_cycles`](#nemo_gym-trajectory-trajectory_record-tool_parent_cycles) | Return only cycle members in the same-instance tool-parent relation. | ### Data [`AgentObservationRecord`](#nemo_gym-trajectory-trajectory_record-AgentObservationRecord) [`NG_AGENT_OBSERVATIONS_KEY`](#nemo_gym-trajectory-trajectory_record-NG_AGENT_OBSERVATIONS_KEY) [`NG_AGENT_SANDBOX_OBSERVATION_KEY`](#nemo_gym-trajectory-trajectory_record-NG_AGENT_SANDBOX_OBSERVATION_KEY) [`NG_TRAJECTORY_KEY`](#nemo_gym-trajectory-trajectory_record-NG_TRAJECTORY_KEY) [`NG_VERIFIER_INVOCATIONS_KEY`](#nemo_gym-trajectory-trajectory_record-NG_VERIFIER_INVOCATIONS_KEY) ### API ```python class nemo_gym.trajectory.trajectory_record.AgentEpisode( response: nemo_gym.core.responses_api_types.NeMoGymResponse, observations: nemo_gym.trajectory.trajectory_record.AgentObservationBundle ) ``` Dataclass An Agent response and the observations available at its execution boundary. **`observations`** `AgentObservationBundle` --- **`response`** `NeMoGymResponse` --- ```python class nemo_gym.trajectory.trajectory_record.AgentInvocation() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) One agent activation or subagent invocation observed by a harness. A harness's saved session may span several invocations; `harness_session_id` retains that context while `invocation_id` owns each activation's evidence. Harness adapters should report an iteration-limit event through an `on_iteration_limit_reached(*, invocation_id: str) -> None` observer hook. The hook records `stop_reason="max_iterations"` once on that invocation, leaving parent invocations, budgets, and summary generation alone. Harnesses with native stop events can call it directly; compatibility shims may call it at the point where the harness detects its limit. **`activation_id`** `Optional[int]` --- **`agent_instance`** `Optional[str] = Field(default=None, min_length=1)` --- **`conversation`** `list[NeMoGymResponseInputItem]` --- **`duration_ms`** `Optional[float] = Field(default=None, ge=0)` --- **`error_type`** `Optional[str] = None` --- **`harness_session_id`** `Optional[str]` --- **`in_session`** `bool | None = Field(default=None, strict=True)` --- **`invocation_id`** `str` --- **`kind`** `Literal['agent_invocation'] = 'agent_invocation'` --- **`model_calls`** `list[ModelCallRef] = Field(default_factory=list)` --- **`parent_invocation_id`** `Optional[str] = None` --- **`spawned_by_tool_call_id`** `Optional[str] = None` --- **`status`** `Literal['completed', 'failed', 'incomplete', 'unknown']` --- **`stop_reason`** `Optional[str]` --- **`verification_skipped`** `bool | None = Field(default=None, strict=True)` --- ```python nemo_gym.trajectory.trajectory_record.AgentInvocation._serialize_optional_metadata( handler: pydantic.SerializerFunctionWrapHandler ) ``` ```python class nemo_gym.trajectory.trajectory_record.AgentObservationBundle() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) Normalized observations returned by one Agent Server for one rollout. **`gaps`** `list[ObservationGap] = Field(default_factory=list)` --- **`records`** `list[AgentObservationRecord]` --- **`schema_version`** `Literal['1.0', '2.0'] = '1.0'` --- **`source`** `str` --- ```python nemo_gym.trajectory.trajectory_record.AgentObservationBundle.decode_version( value: typing.Any ) -> typing.Any ``` classmethod ```python nemo_gym.trajectory.trajectory_record.AgentObservationBundle.serialize_version( handler: pydantic.SerializerFunctionWrapHandler ) ``` ```python nemo_gym.trajectory.trajectory_record.AgentObservationBundle.validate_identity() -> 'AgentObservationBundle' ``` Require unique invocation IDs and reject cycles in the observed parent graph. ```python class nemo_gym.trajectory.trajectory_record.ContextCompactionObservation() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) An explicit context-compaction event reported by the Agent harness. **`after_model_call`** `Optional[ModelCallRef]` --- **`before_model_call`** `Optional[ModelCallRef]` --- **`first_kept_item_id`** `Optional[str] = None` --- **`invocation_id`** `str` --- **`kind`** `Literal['context_compaction'] = 'context_compaction'` --- **`model_calls`** `list[ModelCallRef]` --- **`observed_at`** `Optional[float] = None` --- **`outcome`** `Literal['completed', 'failed', 'aborted', 'unknown'] = 'unknown'` --- **`source_message_ids`** `list[str]` --- **`source_model_ref`** `Optional[ModelServerRef] = None` --- **`summary`** `Optional[str] = None` --- **`tokens_after`** `Optional[int]` --- **`tokens_before`** `Optional[int]` --- **`trigger`** `Optional[str] = None` --- ```python class nemo_gym.trajectory.trajectory_record.GradingOperationObservation() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) A source-reviewed grading calculation actually executed inside a verifier invocation. `llm_judge` grades from judge calls made through `call_judge`; its dependencies are those calls' model and response ids, and it completes only when the reviewed call site marked an explicitly parsed verdict. **`dependencies`** `list[ModelCallRef] | None = None` --- **`kind`** `Literal['programmatic_calculation', 'llm_judge'] = 'programmatic_calculation'` --- **`operation_id`** `str = Field(min_length=1)` --- **`outcome`** `Literal['completed', 'failed', 'unknown'] = 'unknown'` --- **`result_binding_sha256`** `str | None = Field(default=None, pattern='^[0-9a-f]{64}$')` --- ```python class nemo_gym.trajectory.trajectory_record.ModelCallRef() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) Stable identifiers an Agent integration can observe for one model call. **`model_call_id`** `Optional[str] = None` --- **`model_ref`** `Optional[ModelServerRef] = None` --- **`response_id`** `Optional[str] = None` --- ```python nemo_gym.trajectory.trajectory_record.ModelCallRef.validate_join_key() -> 'ModelCallRef' ``` ```python class nemo_gym.trajectory.trajectory_record.ObservationGap() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) A fact that the selected integration could not observe or join exactly. **`agent_instance`** `Optional[str] = None` --- **`code`** `str` --- **`detail`** `Optional[str] = None` --- **`invocation_id`** `Optional[str] = None` --- **`span_id`** `Optional[str] = None` --- ```python nemo_gym.trajectory.trajectory_record.ObservationGap.matches_tool( tool: nemo_gym.trajectory.trajectory_record.ToolCallObservation ) -> bool ``` A supplied owner constrains the span; missing ownership never selects between matches. ```python class nemo_gym.trajectory.trajectory_record.ObservationModel() ``` **Bases:** `BaseModel` **`model_config`** --- ```python class nemo_gym.trajectory.trajectory_record.OutputBudgetRefit() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) A model call the server sent again with its output budget lowered, because the budget it requested exceeded the engine's whole context window. **`applied_max_tokens`** `int = Field(ge=1)` --- **`requested_max_tokens`** `int = Field(ge=1)` --- **`window`** `int = Field(ge=1)` --- ```python class nemo_gym.trajectory.trajectory_record.SandboxObservation() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) Outcome and lifetime resource usage reported by a sandbox-owning harness. **`cleanup_errors`** `list[str]` --- **`cpu_time_s`** `Optional[float]` --- **`error_message`** `Optional[str] = None` --- **`error_type`** `Optional[str] = None` --- **`exit_code`** `Optional[int] = None` --- **`failure_id`** `Optional[str]` --- **`kind`** `Literal['sandbox'] = 'sandbox'` --- **`observed_client_interval_s`** `Optional[float]` --- **`outcome`** `Literal['completed', 'failed', 'timeout', 'oom', 'sandbox_error', 'cancelled', 'unknown'] = 'unknown'` --- **`peak_memory_mib`** `Optional[float]` --- **`provider`** `Optional[str] = None` --- **`resource_usage_source`** `Optional[str] = None` --- **`role`** `Literal['agent', 'verifier', 'environment']` --- **`sandbox_id`** `Optional[str] = None` --- **`stage`** `FailureStage | None` --- **`teardown_confirmed`** `Optional[bool]` --- **`wall_time_s`** `Optional[float] = Field(default=None, ge=0)` --- ```python class nemo_gym.trajectory.trajectory_record.ToolCallObservation() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) Raw tool evidence; span and parent IDs are local to its producer until imported. **`agent_instance`** `Optional[str] = Field(default=None, min_length=1)` --- **`completed_at`** `Optional[float] = None` --- **`duration_ms`** `Optional[float] = Field(default=None, ge=0)` --- **`error_type`** `Optional[str] = None` --- **`failure_id`** `Optional[str] = None` --- **`invocation_id`** `str` --- **`kind`** `Literal['tool_call'] = 'tool_call'` --- **`model_call_id`** `Optional[str] = None` --- **`model_call_ref`** `Optional[ModelCallRef] = None` --- **`native_tool_call_id`** `Optional[str] = None` --- **`parent_span_id`** `Optional[str] = None` --- **`producer_id`** `Optional[str] = Field(default=None, min_length=1)` --- **`sandbox_id`** `Optional[str]` --- **`source_message_id`** `Optional[str] = None` --- **`source_model_ref`** `Optional[ModelServerRef] = None` --- **`span_id`** `Optional[str] = Field(default=None, min_length=1)` --- **`started_at`** `Optional[float] = None` --- **`status`** `Literal['completed', 'failed', 'timeout', 'cancelled', 'incomplete', 'unknown'] = 'unknown'` --- **`timing_source`** `Optional[Literal['executor', 'artifact', 'harness']] = None` --- **`timings`** `list[OperationTiming] = Field(default_factory=list)` --- **`tool_call_id`** `str` --- **`tool_name`** `Optional[str] = None` --- ```python nemo_gym.trajectory.trajectory_record.ToolCallObservation.validate_timing() -> 'ToolCallObservation' ``` ```python class nemo_gym.trajectory.trajectory_record.TrajectoryModelCall() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) **`agent_instance`** `Optional[str] = Field(default=None, min_length=1)` --- **`client_assistant_message_id`** `Optional[str] = None` --- **`client_session_id`** `Optional[str] = None` --- **`clock_domain`** `Optional[str] = Field(default=None, min_length=1)` --- **`completed_at`** `Optional[float] = None` --- **`duration_ms`** `Optional[float] = Field(default=None, ge=0)` --- **`model_call_id`** `Optional[str] = None` --- **`request`** `Optional[Any] = None` --- **`response`** `Optional[Any] = None` --- **`response_metadata`** `TrajectoryResponseMetadata = Field(default_factory=TrajectoryResponseMetadata)` --- **`started_at`** `Optional[float] = None` --- **`token_stats`** `TrajectoryTokenStats = Field(default_factory=TrajectoryTokenStats)` --- ```python class nemo_gym.trajectory.trajectory_record.TrajectoryRecord() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) **`capture_key`** `str` The attempt-qualified episode identity (called rollout_id in v1). --- **`failure_events`** `dict[str, dict[str, Any]] = Field(default_factory=dict)` --- **`gaps`** `list[ObservationGap] = Field(default_factory=list)` --- **`invocations`** `list[AgentInvocation] = Field(default_factory=list)` --- **`model_calls`** `list[TrajectoryModelCall] = Field(default_factory=list)` --- **`rollout_id`** `str` --- **`schema_version`** `Literal['1.0', '2.0'] = '1.0'` --- **`task_id`** `str` --- **`tool_calls`** `list[TrajectoryToolCall] = Field(default_factory=list)` --- **`turns`** `list[TrajectoryTurn] = Field(default_factory=list)` --- **`verifier_invocations`** `list[VerifierInvocation] = Field(default_factory=list)` --- ```python nemo_gym.trajectory.trajectory_record.TrajectoryRecord.decode_identity( value: typing.Any ) -> typing.Any ``` classmethod ```python nemo_gym.trajectory.trajectory_record.TrajectoryRecord.serialize_version( handler: pydantic.SerializerFunctionWrapHandler, info: pydantic.SerializationInfo ) ``` ```python nemo_gym.trajectory.trajectory_record.TrajectoryRecord.validate_identity() -> 'TrajectoryRecord' ``` ```python class nemo_gym.trajectory.trajectory_record.TrajectoryResponseMetadata() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) **`applied_max_tokens`** `Optional[int] = Field(default=None, ge=1)` --- **`dialect`** `Optional[str] = None` --- **`error_category`** `Optional[str] = None` --- **`failure_report`** `FailureReport | None = None` --- **`finish_reason`** `Optional[str] = None` --- **`latency_ttft_ms`** `Optional[float] = Field(default=None, ge=0)` --- **`local_response_reason`** `Optional[str] = None` --- **`model`** `Optional[str] = None` --- **`model_operations`** `list[ModelOperationObservation] | None = None` --- **`model_ref`** `Optional[ModelServerRef] = None` --- **`output_budget_refit`** `Optional[OutputBudgetRefit] = None` --- **`requested_max_tokens`** `Optional[int] = Field(default=None, ge=1)` --- **`response_id`** `Optional[str] = None` --- **`response_source`** `Optional[Literal['upstream', 'local']] = None` --- **`response_status`** `Optional[str] = None` --- **`status_code`** `Optional[int] = None` --- **`upstream_attempted`** `Optional[bool] = None` --- **`upstream_status_code`** `Optional[int] = None` --- ```python class nemo_gym.trajectory.trajectory_record.TrajectoryTokenStats() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) **`cached_tokens`** `Optional[int] = Field(default=None, ge=0)` --- **`completion_tokens`** `Optional[int] = Field(default=None, ge=0)` --- **`prompt_tokens`** `Optional[int] = Field(default=None, ge=0)` --- **`reasoning_tokens`** `Optional[int] = Field(default=None, ge=0)` --- **`total_tokens`** `Optional[int] = Field(default=None, ge=0)` --- ```python class nemo_gym.trajectory.trajectory_record.TrajectoryToolCall() ``` **Bases:** [ToolCallObservation](#nemo_gym-trajectory-trajectory_record-ToolCallObservation) Canonical tool attempt with qualified span IDs and model-visible output. Unlike raw ToolCallObservation inputs, these IDs have already crossed the import boundary and must remain unchanged on replay. **`output`** `Optional[Any] = None` --- ```python class nemo_gym.trajectory.trajectory_record.TrajectoryTurn() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) **`activation_id`** `Optional[int]` --- **`agent_instance`** `Optional[str] = Field(default=None, min_length=1)` --- **`answer`** `Optional[Any] = None` --- **`invocation_id`** `str` --- **`model_calls`** `list[ModelCallRef] = Field(default_factory=list)` --- **`question`** `Optional[Any] = None` --- **`reasoning_content`** `Optional[Any] = None` --- **`resolved`** `Optional[bool] = None` --- **`rollout_id`** `str` --- **`source_message_id`** `Optional[str]` --- **`source_model_ref`** `Optional[ModelServerRef]` --- **`step_count`** `int` --- **`task_id`** `str` --- **`timestamp`** `float` --- **`turn_no`** `int` --- ```python nemo_gym.trajectory.trajectory_record.TrajectoryTurn._serialize_optional_metadata( handler: pydantic.SerializerFunctionWrapHandler ) ``` ```python class nemo_gym.trajectory.trajectory_record.VerifierGradingRecord() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) The grade a verifier returned without asking any model, recorded by the shared verify path. It names the verifier (its resources server), the score and the verifier's verdict on it (`masked` when the verifier marked the score as no valid measurement), and binds the graded response and score like a grading operation. A verify execution that sent a model request records none: its grading is observed only through its captured calls. **`result_binding_sha256`** `str = Field(pattern='^[0-9a-f]{64}$')` --- **`score`** `float` --- **`verdict`** `Literal['scored', 'masked']` --- **`verifier`** `str = Field(min_length=1)` --- ```python class nemo_gym.trajectory.trajectory_record.VerifierInvocation() ``` **Bases:** [ObservationModel](#nemo_gym-trajectory-trajectory_record-ObservationModel) One verify() execution, its outcome, and all model attempts made while grading. Captured calls establish ownership. The shared verify boundary records the execution outcome, including for programmatic verifiers with no model calls. **`failure_kind`** `Optional[str] = None` --- **`grading_operations`** `list[GradingOperationObservation] | None` --- **`grading_record`** `VerifierGradingRecord | None` --- **`invocation_id`** `str` --- **`kind`** `Literal['verifier_invocation'] = 'verifier_invocation'` --- **`model_calls`** `list[ModelCallRef] = Field(default_factory=list)` --- **`sandbox_observations`** `list[SandboxObservation] | None` --- **`status`** `Literal['completed', 'failed', 'incomplete', 'unknown'] = 'unknown'` --- ```python nemo_gym.trajectory.trajectory_record.VerifierInvocation.validate_failure_kind( value: typing.Optional[str] ) -> typing.Optional[str] ``` classmethod ```python nemo_gym.trajectory.trajectory_record.exclude_verifier_model_calls( calls: collections.abc.Iterable[nemo_gym.trajectory.model_call_record.ModelCallRecord], verifier_invocations: collections.abc.Iterable[nemo_gym.trajectory.trajectory_record.VerifierInvocation] ) -> list[nemo_gym.trajectory.model_call_record.ModelCallRecord] ``` Exclude explicitly verifier-owned attempts from agent and step joins. ```python nemo_gym.trajectory.trajectory_record.failure_observations( value: typing.Any ) -> dict[str, typing.Any] ``` Keep the existing typed observation slots across an unsuccessful HTTP reply. Error strings are not searched recursively: only producer-owned top-level fields are accepted, and malformed observations remain unavailable. ```python nemo_gym.trajectory.trajectory_record.grading_result_binding( result: object ) -> str | None ``` Bind the actual graded response and numeric reward/components, without exporting their payload. ```python nemo_gym.trajectory.trajectory_record.join_agent_model_calls( bundle: nemo_gym.trajectory.trajectory_record.AgentObservationBundle, calls: collections.abc.Iterable[nemo_gym.trajectory.model_call_record.ModelCallRecord] ) -> nemo_gym.trajectory.trajectory_record.AgentObservationBundle ``` Join captured calls to agent invocations and their compactions without guessing ownership. ```python nemo_gym.trajectory.trajectory_record.join_assistant_message_calls( bundle: nemo_gym.trajectory.trajectory_record.AgentObservationBundle, trajectory: nemo_gym.trajectory.trajectory_record.TrajectoryRecord | None, calls: list['ModelCallRecord'] ) -> tuple[nemo_gym.trajectory.trajectory_record.AgentObservationBundle, nemo_gym.trajectory.trajectory_record.TrajectoryRecord | None] ``` Join persisted messages to HTTP attempts using session, message and model-server IDs. Run after invocation ownership resolution. Summary attempts belong to compactions; calls without a saved message (for example title generation) stay invocation-owned. Producers declare source_message_id(s), source_model_ref and invocation IDs. Source labels are not dispatch keys. No content or timing fallback is used. ```python nemo_gym.trajectory.trajectory_record.tool_parent_cycles( tools: collections.abc.Iterable[nemo_gym.trajectory.trajectory_record.ToolCallObservation] ) -> set[tuple[str | None, str]] ``` Return only cycle members in the same-instance tool-parent relation. ```python nemo_gym.trajectory.trajectory_record.AgentObservationRecord = Annotated[AgentInvocation | ToolCallObservation | ContextCompactionObservation |... ``` ```python nemo_gym.trajectory.trajectory_record.NG_AGENT_OBSERVATIONS_KEY = 'ng_agent_observations' ``` ```python nemo_gym.trajectory.trajectory_record.NG_AGENT_SANDBOX_OBSERVATION_KEY = 'ng_agent_sandbox_observation' ``` ```python nemo_gym.trajectory.trajectory_record.NG_TRAJECTORY_KEY = 'ng_trajectory' ``` ```python nemo_gym.trajectory.trajectory_record.NG_VERIFIER_INVOCATIONS_KEY = 'ng_verifier_invocations' ```