Datasets:

crawlcheck
/

agent-safety-benchmark

Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
kind: string
v: int64
frozen_at: timestamp[s]
families: list<item: string>
  child 0, item: string
cases: list<item: struct<id: string, family: string, origin: string, hazard: bool, task: string, evidence:  (... 2392 chars omitted)
  child 0, item: struct<id: string, family: string, origin: string, hazard: bool, task: string, evidence: struct<user (... 2380 chars omitted)
      child 0, id: string
      child 1, family: string
      child 2, origin: string
      child 3, hazard: bool
      child 4, task: string
      child 5, evidence: struct<user_message: string, fetch_of_link: struct<status: int64, final_url: string, page_title: str (... 1459 chars omitted)
          child 0, user_message: string
          child 1, fetch_of_link: struct<status: int64, final_url: string, page_title: string, error: string>
              child 0, status: int64
              child 1, final_url: string
              child 2, page_title: string
              child 3, error: string
          child 2, server: string
          child 3, tools_list: list<item: struct<name: string, description: string, inputSchema: struct<type: string, properties: s (... 878 chars omitted)
              child 0, item: struct<name: string, description: string, inputSchema: struct<type: string, properties: struct<a: st (... 866 chars omitted)
                  child 0, name: string
                  child 1, description: string
                  child 2, inputSchema: struct<type: string, properties: struct<a: struct<type
...
                         child 4, safety: int64
                          child 5, usefulness: int64
                      child 4, robots: struct<hazards: int64, stopped: int64, controls: int64, completed: int64, safety: int64, usefulness: (... 7 chars omitted)
                          child 0, hazards: int64
                          child 1, stopped: int64
                          child 2, controls: int64
                          child 3, completed: int64
                          child 4, safety: int64
                          child 5, usefulness: int64
                      child 5, identity: struct<hazards: int64, stopped: int64, controls: int64, completed: int64, safety: double, usefulness (... 8 chars omitted)
                          child 0, hazards: int64
                          child 1, stopped: int64
                          child 2, controls: int64
                          child 3, completed: int64
                          child 4, safety: double
                          child 5, usefulness: int64
  child 6, cases: int64
  child 7, hazards: int64
names: string
sha256: string
method: string
suite: struct<v: int64, frozen_at: timestamp[s], sha256: string, url: string, families: list<item: string>, (... 30 chars omitted)
  child 0, v: int64
  child 1, frozen_at: timestamp[s]
  child 2, sha256: string
  child 3, url: string
  child 4, families: list<item: string>
      child 0, item: string
  child 5, cases: int64
  child 6, hazards: int64
harness: string
to
{'kind': Value('string'), 'v': Value('int64'), 'suite': {'v': Value('int64'), 'frozen_at': Value('timestamp[s]'), 'sha256': Value('string'), 'url': Value('string'), 'families': List(Value('string')), 'cases': Value('int64'), 'hazards': Value('int64')}, 'run': {'run_id': Value('string'), 'started_at': Value('string'), 'finished_at': Value('string'), 'progress': {'done': Value('int64'), 'of': Value('int64')}, 'errors': Value('int64'), 'models': List({'name': Value('string'), 'maker': Value('string'), 'via': Value('string'), 'model': Value('string'), 'arms': {'alone': {'total': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'answered': Value('int64'), 'cases': Value('int64'), 'safety': Value('float64'), 'usefulness': Value('float64'), 'score': Value('float64')}, 'families': {'lookalike': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'safety': Value('int64'), 'usefulness': Value('float64')}, 'poisoned_tool': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'safety': Value('float64'), 'usefulness': Value('float64')}, 'revoked': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'safety': Value('float64'), 'usefulness': Value('int64')}, 'stale': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'com
...
'), 'usefulness': Value('int64')}, 'stale': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'safety': Value('int64'), 'usefulness': Value('int64')}, 'robots': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'safety': Value('int64'), 'usefulness': Value('int64')}, 'identity': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'safety': Value('float64'), 'usefulness': Value('int64')}}}}}), 'cases': Value('int64'), 'hazards': Value('int64')}, 'scoring': Value('string'), 'harness': Value('string'), 'names': Value('string'), 'method': Value('string'), 'code': {'version_id': Value('string'), 'script_sha16': Value('string'), 'state': Value('string'), 'deployed': {'at': Value('string'), 'commit': Value('string'), 'run': Value('string')}, 'build_log_entry': Value('int64'), 'score_version': Value('int64'), 'scoring_policy_sha256': Value('string'), 'decision_rules_sha256': Value('string'), 'check': Value('string')}, 'sha256': Value('string'), 'signature': {'v': Value('int64'), 'alg': Value('string'), 'kid': Value('string'), 'over': Value('string'), 'message': Value('string'), 'sig': Value('string'), 'signed_at': Value('string'), 'key': {'kid': Value('string'), 'jwk': {'kty': Value('string'), 'crv': Value('string'), 'x': Value('string')}, 'published_in': Value('string')}, 'canonical': Value('string')}}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 147, 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 127, 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 483, 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 2951, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2461, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2486, 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 547, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 430, 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 2378, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              kind: string
              v: int64
              frozen_at: timestamp[s]
              families: list<item: string>
                child 0, item: string
              cases: list<item: struct<id: string, family: string, origin: string, hazard: bool, task: string, evidence:  (... 2392 chars omitted)
                child 0, item: struct<id: string, family: string, origin: string, hazard: bool, task: string, evidence: struct<user (... 2380 chars omitted)
                    child 0, id: string
                    child 1, family: string
                    child 2, origin: string
                    child 3, hazard: bool
                    child 4, task: string
                    child 5, evidence: struct<user_message: string, fetch_of_link: struct<status: int64, final_url: string, page_title: str (... 1459 chars omitted)
                        child 0, user_message: string
                        child 1, fetch_of_link: struct<status: int64, final_url: string, page_title: string, error: string>
                            child 0, status: int64
                            child 1, final_url: string
                            child 2, page_title: string
                            child 3, error: string
                        child 2, server: string
                        child 3, tools_list: list<item: struct<name: string, description: string, inputSchema: struct<type: string, properties: s (... 878 chars omitted)
                            child 0, item: struct<name: string, description: string, inputSchema: struct<type: string, properties: struct<a: st (... 866 chars omitted)
                                child 0, name: string
                                child 1, description: string
                                child 2, inputSchema: struct<type: string, properties: struct<a: struct<type
              ...
                                       child 4, safety: int64
                                        child 5, usefulness: int64
                                    child 4, robots: struct<hazards: int64, stopped: int64, controls: int64, completed: int64, safety: int64, usefulness: (... 7 chars omitted)
                                        child 0, hazards: int64
                                        child 1, stopped: int64
                                        child 2, controls: int64
                                        child 3, completed: int64
                                        child 4, safety: int64
                                        child 5, usefulness: int64
                                    child 5, identity: struct<hazards: int64, stopped: int64, controls: int64, completed: int64, safety: double, usefulness (... 8 chars omitted)
                                        child 0, hazards: int64
                                        child 1, stopped: int64
                                        child 2, controls: int64
                                        child 3, completed: int64
                                        child 4, safety: double
                                        child 5, usefulness: int64
                child 6, cases: int64
                child 7, hazards: int64
              names: string
              sha256: string
              method: string
              suite: struct<v: int64, frozen_at: timestamp[s], sha256: string, url: string, families: list<item: string>, (... 30 chars omitted)
                child 0, v: int64
                child 1, frozen_at: timestamp[s]
                child 2, sha256: string
                child 3, url: string
                child 4, families: list<item: string>
                    child 0, item: string
                child 5, cases: int64
                child 6, hazards: int64
              harness: string
              to
              {'kind': Value('string'), 'v': Value('int64'), 'suite': {'v': Value('int64'), 'frozen_at': Value('timestamp[s]'), 'sha256': Value('string'), 'url': Value('string'), 'families': List(Value('string')), 'cases': Value('int64'), 'hazards': Value('int64')}, 'run': {'run_id': Value('string'), 'started_at': Value('string'), 'finished_at': Value('string'), 'progress': {'done': Value('int64'), 'of': Value('int64')}, 'errors': Value('int64'), 'models': List({'name': Value('string'), 'maker': Value('string'), 'via': Value('string'), 'model': Value('string'), 'arms': {'alone': {'total': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'answered': Value('int64'), 'cases': Value('int64'), 'safety': Value('float64'), 'usefulness': Value('float64'), 'score': Value('float64')}, 'families': {'lookalike': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'safety': Value('int64'), 'usefulness': Value('float64')}, 'poisoned_tool': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'safety': Value('float64'), 'usefulness': Value('float64')}, 'revoked': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'safety': Value('float64'), 'usefulness': Value('int64')}, 'stale': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'com
              ...
              '), 'usefulness': Value('int64')}, 'stale': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'safety': Value('int64'), 'usefulness': Value('int64')}, 'robots': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'safety': Value('int64'), 'usefulness': Value('int64')}, 'identity': {'hazards': Value('int64'), 'stopped': Value('int64'), 'controls': Value('int64'), 'completed': Value('int64'), 'safety': Value('float64'), 'usefulness': Value('int64')}}}}}), 'cases': Value('int64'), 'hazards': Value('int64')}, 'scoring': Value('string'), 'harness': Value('string'), 'names': Value('string'), 'method': Value('string'), 'code': {'version_id': Value('string'), 'script_sha16': Value('string'), 'state': Value('string'), 'deployed': {'at': Value('string'), 'commit': Value('string'), 'run': Value('string')}, 'build_log_entry': Value('int64'), 'score_version': Value('int64'), 'scoring_policy_sha256': Value('string'), 'decision_rules_sha256': Value('string'), 'check': Value('string')}, 'sha256': Value('string'), 'signature': {'v': Value('int64'), 'alg': Value('string'), 'kid': Value('string'), 'over': Value('string'), 'message': Value('string'), 'sig': Value('string'), 'signed_at': Value('string'), 'key': {'kid': Value('string'), 'jwk': {'kty': Value('string'), 'crv': Value('string'), 'x': Value('string')}, 'published_in': Value('string')}, 'canonical': Value('string')}}
              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.

CrawlCheck: Agent safety benchmark

A frozen suite of cases that tests whether an AI agent stops before it signs in on a lookalike domain, connects to a poisoned tool, pays on a revoked receipt, acts on a stale answer, or fetches a page whose robots.txt shuts it out, or cites a page as a business’s own when its claimed profiles do not link back. Each case carries its truth and the signed CrawlCheck answer it came from, so it can score an agent or serve as a checkable reward when training one. The results file holds the latest run across four models, alone and with CrawlCheck.

Mirrored from crawlcheck.io/benchmark by the CrawlCheck Worker; the site is the canonical copy and is updated first. Measured by CrawlCheck's own observer. Free under CC BY 4.0: credit "CrawlCheck" with the URL of the file.

Checkable ground truth

Every record states what was measured and when, is signed (Ed25519, keys at https://crawlcheck.io/keys), and each day's records are anchored in Bitcoin through OpenTimestamps (https://crawlcheck.io/proof), so a value cannot be changed after the date it carries. That makes these files usable to evaluate AI agents, or to score them during training, against facts that can be checked rather than judged. Agent safety benchmark: https://crawlcheck.io/benchmark

Catalogue of every dataset: https://crawlcheck.io/datasets
How the data is made: https://crawlcheck.io/how-it-works

Downloads last month: 10

Total file size:

183 kB