Datasets:

quotidianant
/

tracejudge-traces

Tags:

License:

Dataset card Files Files and versions

xet

Community

Dataset Viewer

Duplicate

Split (1)

train

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
schema_version: string
trajectory_id: string
session_id: string
agent: struct<name: string, extra: struct<task_id: string, tracejudge_task_id: string, domain: string, sour (... 55 chars omitted)
  child 0, name: string
  child 1, extra: struct<task_id: string, tracejudge_task_id: string, domain: string, source_dataset: string, benchmar (... 26 chars omitted)
      child 0, task_id: string
      child 1, tracejudge_task_id: string
      child 2, domain: string
      child 3, source_dataset: string
      child 4, benchmark: string
      child 5, resolved: bool
steps: list<item: struct<step_id: int64, timestamp: string, source: string, message: string, tool_calls: li (... 765 chars omitted)
  child 0, item: struct<step_id: int64, timestamp: string, source: string, message: string, tool_calls: list<item: st (... 753 chars omitted)
      child 0, step_id: int64
      child 1, timestamp: string
      child 2, source: string
      child 3, message: string
      child 4, tool_calls: list<item: struct<tool_call_id: string, function_name: string, arguments: struct<data_path: string,  (... 51 chars omitted)
          child 0, item: struct<tool_call_id: string, function_name: string, arguments: struct<data_path: string, info_type:  (... 39 chars omitted)
              child 0, tool_call_id: string
              child 1, function_name: string
              child 2, arguments: struct<data_path: string, info_type: string, keyterm: string, code: string>
                  child 0, data_path: 
...
d)
      child 0, node_id: string
      child 1, merged_state: struct<capability: string, intent: string, sub_intent: string, effect: string, target: string, repre (... 249 chars omitted)
          child 0, capability: string
          child 1, intent: string
          child 2, sub_intent: string
          child 3, effect: string
          child 4, target: string
          child 5, representative_input_signature: string
          child 6, member_refs: list<item: struct<trace_id: string, span_id: string, trajectory_position: int64, input_signature: st (... 32 chars omitted)
              child 0, item: struct<trace_id: string, span_id: string, trajectory_position: int64, input_signature: string, resul (... 20 chars omitted)
                  child 0, trace_id: string
                  child 1, span_id: string
                  child 2, trajectory_position: int64
                  child 3, input_signature: string
                  child 4, result_signature: string
          child 7, equivalence_tier: string
          child 8, confidence: double
          child 9, variant_count: int64
      child 2, support: int64
      child 3, pruned: bool
      child 4, intent_stage: string
prune_threshold: int64
num_trajectories: int64
task_id: string
edges: list<item: struct<from_node: string, to_node: string, weight: int64>>
  child 0, item: struct<from_node: string, to_node: string, weight: int64>
      child 0, from_node: string
      child 1, to_node: string
      child 2, weight: int64
to
{'task_id': Value('string'), 'num_trajectories': Value('int64'), 'prune_threshold': Value('int64'), 'nodes': List({'node_id': Value('string'), 'merged_state': {'capability': Value('string'), 'intent': Value('string'), 'sub_intent': Value('string'), 'effect': Value('string'), 'target': Value('string'), 'representative_input_signature': Value('string'), 'member_refs': List({'trace_id': Value('string'), 'span_id': Value('string'), 'trajectory_position': Value('int64'), 'input_signature': Value('string'), 'result_signature': Value('string')}), 'equivalence_tier': Value('string'), 'confidence': Value('float64'), 'variant_count': Value('int64')}, 'support': Value('int64'), 'pruned': Value('bool'), 'intent_stage': Value('string')}), 'edges': List({'from_node': Value('string'), 'to_node': Value('string'), 'weight': Value('int64')})}
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
              schema_version: string
              trajectory_id: string
              session_id: string
              agent: struct<name: string, extra: struct<task_id: string, tracejudge_task_id: string, domain: string, sour (... 55 chars omitted)
                child 0, name: string
                child 1, extra: struct<task_id: string, tracejudge_task_id: string, domain: string, source_dataset: string, benchmar (... 26 chars omitted)
                    child 0, task_id: string
                    child 1, tracejudge_task_id: string
                    child 2, domain: string
                    child 3, source_dataset: string
                    child 4, benchmark: string
                    child 5, resolved: bool
              steps: list<item: struct<step_id: int64, timestamp: string, source: string, message: string, tool_calls: li (... 765 chars omitted)
                child 0, item: struct<step_id: int64, timestamp: string, source: string, message: string, tool_calls: list<item: st (... 753 chars omitted)
                    child 0, step_id: int64
                    child 1, timestamp: string
                    child 2, source: string
                    child 3, message: string
                    child 4, tool_calls: list<item: struct<tool_call_id: string, function_name: string, arguments: struct<data_path: string,  (... 51 chars omitted)
                        child 0, item: struct<tool_call_id: string, function_name: string, arguments: struct<data_path: string, info_type:  (... 39 chars omitted)
                            child 0, tool_call_id: string
                            child 1, function_name: string
                            child 2, arguments: struct<data_path: string, info_type: string, keyterm: string, code: string>
                                child 0, data_path: 
              ...
              d)
                    child 0, node_id: string
                    child 1, merged_state: struct<capability: string, intent: string, sub_intent: string, effect: string, target: string, repre (... 249 chars omitted)
                        child 0, capability: string
                        child 1, intent: string
                        child 2, sub_intent: string
                        child 3, effect: string
                        child 4, target: string
                        child 5, representative_input_signature: string
                        child 6, member_refs: list<item: struct<trace_id: string, span_id: string, trajectory_position: int64, input_signature: st (... 32 chars omitted)
                            child 0, item: struct<trace_id: string, span_id: string, trajectory_position: int64, input_signature: string, resul (... 20 chars omitted)
                                child 0, trace_id: string
                                child 1, span_id: string
                                child 2, trajectory_position: int64
                                child 3, input_signature: string
                                child 4, result_signature: string
                        child 7, equivalence_tier: string
                        child 8, confidence: double
                        child 9, variant_count: int64
                    child 2, support: int64
                    child 3, pruned: bool
                    child 4, intent_stage: string
              prune_threshold: int64
              num_trajectories: int64
              task_id: string
              edges: list<item: struct<from_node: string, to_node: string, weight: int64>>
                child 0, item: struct<from_node: string, to_node: string, weight: int64>
                    child 0, from_node: string
                    child 1, to_node: string
                    child 2, weight: int64
              to
              {'task_id': Value('string'), 'num_trajectories': Value('int64'), 'prune_threshold': Value('int64'), 'nodes': List({'node_id': Value('string'), 'merged_state': {'capability': Value('string'), 'intent': Value('string'), 'sub_intent': Value('string'), 'effect': Value('string'), 'target': Value('string'), 'representative_input_signature': Value('string'), 'member_refs': List({'trace_id': Value('string'), 'span_id': Value('string'), 'trajectory_position': Value('int64'), 'input_signature': Value('string'), 'result_signature': Value('string')}), 'equivalence_tier': Value('string'), 'confidence': Value('float64'), 'variant_count': Value('int64')}, 'support': Value('int64'), 'pruned': Value('bool'), 'intent_stage': Value('string')}), 'edges': List({'from_node': Value('string'), 'to_node': Value('string'), 'weight': Value('int64')})}
              because column names don't match

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

TraceJudge community agent traces

Community-submitted agent execution traces, normalized to ATIF (Agent Trajectory Interchange Format, ATIF-v1.7 — Harbor Framework). Traces come from any agent with a supported adapter (Claude Code, Codex, OpenHands, OpenAI chat-messages) and are validated before upload — every file passes validate_atif() from TraceJudge.

Layout

traces/<agent>/<trajectory_id>.atif.json

One JSON file per trajectory: a flat list of user/agent/system steps with tool calls, observations, reasoning, and token metrics.

Submitting your own traces

git clone https://github.com/fleet-fellowship/TraceJudge && cd TraceJudge
uv venv --python 3.12 .venv && uv pip install --python .venv -e .
hf auth login   # your own HF account

# from a local Claude Code / Codex session
.venv/bin/python scripts/submit_trace.py --list-sessions
.venv/bin/python scripts/submit_trace.py --session 0 --push

# or from a raw trace file (any supported format)
.venv/bin/python scripts/submit_trace.py --trace my_trace.json --push

If you don't have write access, the upload automatically opens a Hub pull request. Traces can contain working directories, file contents, shell commands, and prompts — review before sharing.

Browsing

.venv/bin/python scripts/submit_trace.py --pull     # download everything
.venv/bin/python scripts/view_atif.py               # local web viewer

Downloads last month: 14

Total file size:

2.01 MB