Solstice Agent Reliability Lab
Collection
Open datasets, benchmarks, and browser tools for inspecting, replaying, evaluating, and securing AI-agent trajectories. • 6 items • Updated
Dataset Viewer
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
checkpoint_id: string
event_id: string
state: struct<phase: string>
child 0, phase: string
trace_id: string
comparison_classification: string
events: list<item: struct<actor: string, event_id: string, event_type: string, output_summary: string, paren (... 93 chars omitted)
child 0, item: struct<actor: string, event_id: string, event_type: string, output_summary: string, parent_event_id: (... 81 chars omitted)
child 0, actor: string
child 1, event_id: string
child 2, event_type: string
child 3, output_summary: string
child 4, parent_event_id: string
child 5, sequence: int64
child 6, state_delta: struct<control: string>
child 0, control: string
child 7, timestamp: int64
branch_id: string
final_outcome: struct<policy_compliance: bool, steps: int64, synthetic_latency_ms: int64, task_success: bool>
child 0, policy_compliance: bool
child 1, steps: int64
child 2, synthetic_latency_ms: int64
child 3, task_success: bool
parent_checkpoint_id: string
safety_findings: list<item: null>
child 0, item: null
final_state: struct<control: string>
child 0, control: string
to
{'branch_id': Value('string'), 'comparison_classification': Value('string'), 'events': List({'actor': Value('string'), 'event_id': Value('string'), 'event_type': Value('string'), 'output_summary': Value('string'), 'parent_event_id': Value('string'), 'sequence': Value('int64'), 'state_delta': {'control': Value('string')}, 'timestamp': Value('int64')}), 'final_outcome': {'policy_compliance': Value('bool'), 'steps': Value('int64'), 'synthetic_latency_ms': Value('int64'), 'task_success': Value('bool')}, 'final_state': {'control': Value('string')}, 'parent_checkpoint_id': Value('string'), 'safety_findings': List(Value('null')), 'trace_id': Value('string')}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 149, in get_rows_or_raise
return get_rows(
dataset=dataset,
...<4 lines>...
column_names=column_names,
)
File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
return func(*args, **kwargs)
File "/src/services/worker/src/worker/utils.py", line 129, in get_rows
rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
File "/src/services/worker/src/worker/utils.py", line 489, in safe_iter
yield from ds.decode(False) if ds.features else ds
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2818, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2355, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2380, in _iter_arrow
for key, pa_table in self.ex_iterable._iter_arrow():
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 536, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, in _iter_arrow
for key, pa_table in self.generate_tables_fn(**gen_kwags):
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2369, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2297, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
checkpoint_id: string
event_id: string
state: struct<phase: string>
child 0, phase: string
trace_id: string
comparison_classification: string
events: list<item: struct<actor: string, event_id: string, event_type: string, output_summary: string, paren (... 93 chars omitted)
child 0, item: struct<actor: string, event_id: string, event_type: string, output_summary: string, parent_event_id: (... 81 chars omitted)
child 0, actor: string
child 1, event_id: string
child 2, event_type: string
child 3, output_summary: string
child 4, parent_event_id: string
child 5, sequence: int64
child 6, state_delta: struct<control: string>
child 0, control: string
child 7, timestamp: int64
branch_id: string
final_outcome: struct<policy_compliance: bool, steps: int64, synthetic_latency_ms: int64, task_success: bool>
child 0, policy_compliance: bool
child 1, steps: int64
child 2, synthetic_latency_ms: int64
child 3, task_success: bool
parent_checkpoint_id: string
safety_findings: list<item: null>
child 0, item: null
final_state: struct<control: string>
child 0, control: string
to
{'branch_id': Value('string'), 'comparison_classification': Value('string'), 'events': List({'actor': Value('string'), 'event_id': Value('string'), 'event_type': Value('string'), 'output_summary': Value('string'), 'parent_event_id': Value('string'), 'sequence': Value('int64'), 'state_delta': {'control': Value('string')}, 'timestamp': Value('int64')}), 'final_outcome': {'policy_compliance': Value('bool'), 'steps': Value('int64'), 'synthetic_latency_ms': Value('int64'), 'task_success': Value('bool')}, 'final_state': {'control': Value('string')}, 'parent_checkpoint_id': Value('string'), 'safety_findings': List(Value('null')), 'trace_id': Value('string')}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
DeltaStore Replay Traces
Ten deterministic, synthetic agent trajectories with checkpoints and counterfactual branches. The dataset is for replay tooling and methodology validation, not real-world detector accuracy or model ranking.
Files
traces.jsonl— self-contained canonical traces.branches.jsonl— normalized branch records.checkpoints.jsonl— normalized checkpoint records.comparisons.jsonl— deterministic original-versus-branch comparisons.
All traces are synthetic. No private chain-of-thought, credentials, customer data, or production traces are included. Synthetic latency is fixture metadata, not production performance. A branch is a controlled counterfactual and does not prove what a different model would always do.
Schema: solstice-agent-trace-exchange/v1 is included as
solstice-agent-trace-exchange-v1.schema.json.
Link: DeltaStore Static Space.
Related public tools:
The exchange contract is documented. The Atlas browser app currently accepts STS/normalized JSONL rather than the complete DeltaStore exchange envelope.
- Downloads last month
- 13