Dataset Viewer
Auto-converted to Parquet Duplicate
instance_id
stringlengths
62
76
benchmark_name
stringclasses
1 value
benchmark_version
stringclasses
1 value
dataset
stringclasses
1 value
task_type
stringclasses
1 value
prompt
stringclasses
7 values
answer_schema
dict
input_jsonl
stringlengths
1.54M
2.29M
input_artifact
dict
evaluation
dict
reasoning_family
stringclasses
7 values
difficulty_tags
listlengths
5
5
candidate_count
int64
2
74
distractor_count
int64
17
868
flaws_cloudtrail.objective_hard_v3.apigateway_restapi_event_profile.000
CybersecJSONSchemaBench
3.0.0-cloudtrail-objective-hard
flaws_cloudtrail
multi_hop_record_answer
"Group API Gateway records by principal ARN and requestParameters.restApiId, using *account-scope* w(...TRUNCATED)
{"$schema":"https://json-schema.org/draft/2020-12/schema","$id":"https://jsonsecuritybench.local/sch(...TRUNCATED)
"{\"benchmark_source\": \"flaws_cloudtrail\", \"source_file\": \"flaws_cloudtrail_logs/flaws_cloudtr(...TRUNCATED)
{"path":"artifacts/cybersec_jsonschemabench_demo/raw/flaws_cloudtrail_logs.tar","record_count":880,"(...TRUNCATED)
{"primary_metric":"schema_validity_gated_ordered_id_exact","score_fields":["answer_record_ids"],"fie(...TRUNCATED)
apigateway_restapi_event_profile
["apigateway","request_parameter_join","error_reasoning","objective_tiebreaks","long_context_880_rec(...TRUNCATED)
3
350
flaws_cloudtrail.objective_hard_v3.apigateway_restapi_event_profile.001
CybersecJSONSchemaBench
3.0.0-cloudtrail-objective-hard
flaws_cloudtrail
multi_hop_record_answer
"Group API Gateway records by principal ARN and requestParameters.restApiId, using *account-scope* w(...TRUNCATED)
{"$schema":"https://json-schema.org/draft/2020-12/schema","$id":"https://jsonsecuritybench.local/sch(...TRUNCATED)
"{\"benchmark_source\": \"flaws_cloudtrail\", \"source_file\": \"flaws_cloudtrail_logs/flaws_cloudtr(...TRUNCATED)
{"path":"artifacts/cybersec_jsonschemabench_demo/raw/flaws_cloudtrail_logs.tar","record_count":880,"(...TRUNCATED)
{"primary_metric":"schema_validity_gated_ordered_id_exact","score_fields":["answer_record_ids"],"fie(...TRUNCATED)
apigateway_restapi_event_profile
["apigateway","request_parameter_join","error_reasoning","objective_tiebreaks","long_context_880_rec(...TRUNCATED)
3
317
flaws_cloudtrail.objective_hard_v3.apigateway_restapi_event_profile.002
CybersecJSONSchemaBench
3.0.0-cloudtrail-objective-hard
flaws_cloudtrail
multi_hop_record_answer
"Group API Gateway records by principal ARN and requestParameters.restApiId, using *account-scope* w(...TRUNCATED)
{"$schema":"https://json-schema.org/draft/2020-12/schema","$id":"https://jsonsecuritybench.local/sch(...TRUNCATED)
"{\"benchmark_source\": \"flaws_cloudtrail\", \"source_file\": \"flaws_cloudtrail_logs/flaws_cloudtr(...TRUNCATED)
{"path":"artifacts/cybersec_jsonschemabench_demo/raw/flaws_cloudtrail_logs.tar","record_count":880,"(...TRUNCATED)
{"primary_metric":"schema_validity_gated_ordered_id_exact","score_fields":["answer_record_ids"],"fie(...TRUNCATED)
apigateway_restapi_event_profile
["apigateway","request_parameter_join","error_reasoning","objective_tiebreaks","long_context_880_rec(...TRUNCATED)
3
39
flaws_cloudtrail.objective_hard_v3.apigateway_restapi_event_profile.003
CybersecJSONSchemaBench
3.0.0-cloudtrail-objective-hard
flaws_cloudtrail
multi_hop_record_answer
"Group API Gateway records by principal ARN and requestParameters.restApiId, using *account-scope* w(...TRUNCATED)
{"$schema":"https://json-schema.org/draft/2020-12/schema","$id":"https://jsonsecuritybench.local/sch(...TRUNCATED)
"{\"benchmark_source\": \"flaws_cloudtrail\", \"source_file\": \"flaws_cloudtrail_logs/flaws_cloudtr(...TRUNCATED)
{"path":"artifacts/cybersec_jsonschemabench_demo/raw/flaws_cloudtrail_logs.tar","record_count":880,"(...TRUNCATED)
{"primary_metric":"schema_validity_gated_ordered_id_exact","score_fields":["answer_record_ids"],"fie(...TRUNCATED)
apigateway_restapi_event_profile
["apigateway","request_parameter_join","error_reasoning","objective_tiebreaks","long_context_880_rec(...TRUNCATED)
3
30
flaws_cloudtrail.objective_hard_v3.apigateway_restapi_event_profile.004
CybersecJSONSchemaBench
3.0.0-cloudtrail-objective-hard
flaws_cloudtrail
multi_hop_record_answer
"Group API Gateway records by principal ARN and requestParameters.restApiId, using *account-scope* w(...TRUNCATED)
{"$schema":"https://json-schema.org/draft/2020-12/schema","$id":"https://jsonsecuritybench.local/sch(...TRUNCATED)
"{\"benchmark_source\": \"flaws_cloudtrail\", \"source_file\": \"flaws_cloudtrail_logs/flaws_cloudtr(...TRUNCATED)
{"path":"artifacts/cybersec_jsonschemabench_demo/raw/flaws_cloudtrail_logs.tar","record_count":880,"(...TRUNCATED)
{"primary_metric":"schema_validity_gated_ordered_id_exact","score_fields":["answer_record_ids"],"fie(...TRUNCATED)
apigateway_restapi_event_profile
["apigateway","request_parameter_join","error_reasoning","objective_tiebreaks","long_context_880_rec(...TRUNCATED)
3
27
flaws_cloudtrail.objective_hard_v3.apigateway_restapi_event_profile.005
CybersecJSONSchemaBench
3.0.0-cloudtrail-objective-hard
flaws_cloudtrail
multi_hop_record_answer
"Group API Gateway records by principal ARN and requestParameters.restApiId, using *account-scope* w(...TRUNCATED)
{"$schema":"https://json-schema.org/draft/2020-12/schema","$id":"https://jsonsecuritybench.local/sch(...TRUNCATED)
"{\"benchmark_source\": \"flaws_cloudtrail\", \"source_file\": \"flaws_cloudtrail_logs/flaws_cloudtr(...TRUNCATED)
{"path":"artifacts/cybersec_jsonschemabench_demo/raw/flaws_cloudtrail_logs.tar","record_count":880,"(...TRUNCATED)
{"primary_metric":"schema_validity_gated_ordered_id_exact","score_fields":["answer_record_ids"],"fie(...TRUNCATED)
apigateway_restapi_event_profile
["apigateway","request_parameter_join","error_reasoning","objective_tiebreaks","long_context_880_rec(...TRUNCATED)
2
258
flaws_cloudtrail.objective_hard_v3.apigateway_restapi_event_profile.006
CybersecJSONSchemaBench
3.0.0-cloudtrail-objective-hard
flaws_cloudtrail
multi_hop_record_answer
"Group API Gateway records by principal ARN and requestParameters.restApiId, using *account-scope* w(...TRUNCATED)
{"$schema":"https://json-schema.org/draft/2020-12/schema","$id":"https://jsonsecuritybench.local/sch(...TRUNCATED)
"{\"benchmark_source\": \"flaws_cloudtrail\", \"source_file\": \"flaws_cloudtrail_logs/flaws_cloudtr(...TRUNCATED)
{"path":"artifacts/cybersec_jsonschemabench_demo/raw/flaws_cloudtrail_logs.tar","record_count":880,"(...TRUNCATED)
{"primary_metric":"schema_validity_gated_ordered_id_exact","score_fields":["answer_record_ids"],"fie(...TRUNCATED)
apigateway_restapi_event_profile
["apigateway","request_parameter_join","error_reasoning","objective_tiebreaks","long_context_880_rec(...TRUNCATED)
2
59
flaws_cloudtrail.objective_hard_v3.apigateway_restapi_event_profile.007
CybersecJSONSchemaBench
3.0.0-cloudtrail-objective-hard
flaws_cloudtrail
multi_hop_record_answer
"Group API Gateway records by principal ARN and requestParameters.restApiId, using *account-scope* w(...TRUNCATED)
{"$schema":"https://json-schema.org/draft/2020-12/schema","$id":"https://jsonsecuritybench.local/sch(...TRUNCATED)
"{\"benchmark_source\": \"flaws_cloudtrail\", \"source_file\": \"flaws_cloudtrail_logs/flaws_cloudtr(...TRUNCATED)
{"path":"artifacts/cybersec_jsonschemabench_demo/raw/flaws_cloudtrail_logs.tar","record_count":880,"(...TRUNCATED)
{"primary_metric":"schema_validity_gated_ordered_id_exact","score_fields":["answer_record_ids"],"fie(...TRUNCATED)
apigateway_restapi_event_profile
["apigateway","request_parameter_join","error_reasoning","objective_tiebreaks","long_context_880_rec(...TRUNCATED)
2
40
flaws_cloudtrail.objective_hard_v3.apigateway_restapi_event_profile.008
CybersecJSONSchemaBench
3.0.0-cloudtrail-objective-hard
flaws_cloudtrail
multi_hop_record_answer
"Group API Gateway records by principal ARN and requestParameters.restApiId, using *account-scope* w(...TRUNCATED)
{"$schema":"https://json-schema.org/draft/2020-12/schema","$id":"https://jsonsecuritybench.local/sch(...TRUNCATED)
"{\"benchmark_source\": \"flaws_cloudtrail\", \"source_file\": \"flaws_cloudtrail_logs/flaws_cloudtr(...TRUNCATED)
{"path":"artifacts/cybersec_jsonschemabench_demo/raw/flaws_cloudtrail_logs.tar","record_count":880,"(...TRUNCATED)
{"primary_metric":"schema_validity_gated_ordered_id_exact","score_fields":["answer_record_ids"],"fie(...TRUNCATED)
apigateway_restapi_event_profile
["apigateway","request_parameter_join","error_reasoning","objective_tiebreaks","long_context_880_rec(...TRUNCATED)
2
33
flaws_cloudtrail.objective_hard_v3.apigateway_restapi_event_profile.009
CybersecJSONSchemaBench
3.0.0-cloudtrail-objective-hard
flaws_cloudtrail
multi_hop_record_answer
"Group API Gateway records by principal ARN and requestParameters.restApiId, using *account-scope* w(...TRUNCATED)
{"$schema":"https://json-schema.org/draft/2020-12/schema","$id":"https://jsonsecuritybench.local/sch(...TRUNCATED)
"{\"benchmark_source\": \"flaws_cloudtrail\", \"source_file\": \"flaws_cloudtrail_logs/flaws_cloudtr(...TRUNCATED)
{"path":"artifacts/cybersec_jsonschemabench_demo/raw/flaws_cloudtrail_logs.tar","record_count":880,"(...TRUNCATED)
{"primary_metric":"schema_validity_gated_ordered_id_exact","score_fields":["answer_record_ids"],"fie(...TRUNCATED)
apigateway_restapi_event_profile
["apigateway","request_parameter_join","error_reasoning","objective_tiebreaks","long_context_880_rec(...TRUNCATED)
2
28
End of preview. Expand in Data Studio

CybersecJSONSchemaBench CloudTrail Objective Hard v3

This is a 100-problem objective long-context cybersecurity reasoning subset built from the full flAWS CloudTrail corpus.

Each row contains an objective query prompt, a large CloudTrail JSONL context, and the JSON schema the answer must match. Gold answers are deterministic query results over the serialized slice and are not included in this public export.

Families

  • apigateway_restapi_event_profile: 10
  • ec2_runinstances_error_ladder: 20
  • iam_recon_before_mutation: 5
  • principal_event_region_fanout: 20
  • request_parameter_cross_event_profile: 15
  • s3_bucket_event_profile: 20
  • sts_credential_service_fanout: 10

Columns

  • instance_id: Stable problem identifier.
  • prompt: Objective query with ranking and tie-break rules.
  • answer_schema: JSON schema for the expected answer.
  • input_jsonl: Serialized CloudTrail records for the problem context.
  • evaluation: Scoring metadata.
  • reasoning_family: Query family.
  • difficulty_tags: Intended reasoning requirements.
  • candidate_count: Number of competing candidates under the query.
  • distractor_count: Candidate-related records not in the gold answer.
Downloads last month
54