id stringlengths 11 11 | task stringclasses 18
values | messages stringlengths 1.78k 4.74k | domains stringlengths 48 170 | difficulty stringclasses 4
values | outcome stringclasses 5
values | available_tools stringlengths 42 128 | constraints stringlengths 119 266 | expected_outputs stringclasses 18
values | execution_plan stringclasses 18
values | tool_trace stringlengths 2 1.55k | error stringclasses 9
values | clarification stringclasses 8
values | verification stringclasses 18
values | final_result stringlengths 177 462 | scenario_family stringclasses 18
values | split stringclasses 3
values | synthetic bool 1
class | generator_version stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
atm_0000001 | Design and implement a user-friendly application with a backend, database and verification plan. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Design and implement a user-friendly application with a backend, database and verification plan. Context: The ... | ["backend_engineering","database_management","app_development","software_testing","document_creation","workflow_automation"] | very_complex | success | ["email_client","terminal","file_manager","spreadsheet","database","code_editor"] | ["Use accessible and understandable interfaces.","Respect dependencies between tasks.","Validate calculations before presenting conclusions."] | ["application_design","backend_plan","database_schema","test_report"] | [{"step":1,"action":"Prepare application design","purpose":"Produce and validate application design","depends_on":[]},{"step":2,"action":"Prepare backend plan","purpose":"Produce and validate backend plan","depends_on":[1]},{"step":3,"action":"Prepare database schema","purpose":"Produce and validate database schema","d... | [{"step":1,"action":"Prepare application design","tool":"file_manager","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare backend plan","tool":"terminal","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":... | [{"item":"application_design","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"backend_plan","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"database_schema","check":"Check structure, completeness and consistency.","result":"simu... | {"status":"success","completed_outputs":["application_design","backend_plan","database_schema","test_report"],"message":"The planned outputs are represented as completed in this synthetic example."} | fullstack_delivery | train | true | 1.0.0 | ||
atm_0000002 | Design an API-backed product with clear interfaces, input validation and tests. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Design an API-backed product with clear interfaces, input validation and tests. Context: The available informa... | ["software_testing","app_development","backend_engineering","api_integration"] | advanced | impossible | ["browser","terminal","file_manager","document_editor","database"] | ["Report errors instead of silently ignoring them.","Keep outputs structured and easy to inspect.","Validate calculations before presenting conclusions."] | ["api_contract","backend_design","interface_spec","test_plan"] | [{"step":1,"action":"Prepare api contract","purpose":"Produce and validate api contract","depends_on":[]},{"step":2,"action":"Prepare backend design","purpose":"Produce and validate backend design","depends_on":[1]},{"step":3,"action":"Prepare interface spec","purpose":"Produce and validate interface spec","depends_on"... | [{"step":1,"action":"Prepare api contract","tool":"database","status":"simulated","evidence":"Synthetic trace; no external action was executed."}] | [{"item":"api_contract","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"backend_design","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"interface_spec","check":"Check structure, completeness and consistency.","result":"simulated... | {"status":"impossible","completed_outputs":[],"reason":"The task requires information that has not been supplied.","message":"Explain the limitation and identify a feasible alternative."} | api_product | train | true | 1.0.0 | ||
atm_0000003 | Plan a small software product from requirements through implementation, testing and documentation. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Plan a small software product from requirements through implementation, testing and documentation. Context: Th... | ["document_creation","project_management","software_testing","app_development","backend_engineering","api_integration","database_management","research"] | intermediate | failed | ["document_editor","email_client","python","api_client","file_manager","chart_generator"] | ["Report errors instead of silently ignoring them.","Preserve original source data.","Use accessible and understandable interfaces.","Do not claim that an action succeeded without evidence.","Explain uncertainty and missing information."] | ["requirements","architecture","implementation_plan","test_report","documentation"] | [{"step":1,"action":"Prepare requirements","purpose":"Produce and validate requirements","depends_on":[]},{"step":2,"action":"Prepare architecture","purpose":"Produce and validate architecture","depends_on":[1]},{"step":3,"action":"Prepare implementation plan","purpose":"Produce and validate implementation plan","depen... | [{"step":1,"action":"Prepare requirements","tool":"email_client","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare architecture","tool":"file_manager","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":"P... | {"code":"dependency_error","description":"A required dependency is unavailable."} | [{"item":"requirements","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"architecture","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"implementation_plan","check":"Check structure, completeness and consistency.","result":"simula... | {"status":"failed","completed_outputs":[],"message":"The task could not be completed. The failure is reported explicitly.","next_step":"Investigate the error and obtain any missing authorization or input."} | integrated_product | train | true | 1.0.0 | |
atm_0000004 | Collect source information, create structured documents and prepare organized deliverables. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Collect source information, create structured documents and prepare organized deliverables. Context: The workf... | ["document_creation","file_management","research","email_writing"] | very_complex | success | ["spreadsheet","python","chart_generator","document_editor","api_client","email_client","terminal"] | ["Use accessible and understandable interfaces.","Explain uncertainty and missing information.","Do not claim that an action succeeded without evidence.","Report errors instead of silently ignoring them.","Avoid exposing credentials or private information."] | ["source_notes","main_document","file_index","email_draft"] | [{"step":1,"action":"Prepare source notes","purpose":"Produce and validate source notes","depends_on":[]},{"step":2,"action":"Prepare main document","purpose":"Produce and validate main document","depends_on":[1]},{"step":3,"action":"Prepare file index","purpose":"Produce and validate file index","depends_on":[2]},{"st... | [{"step":1,"action":"Prepare source notes","tool":"python","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare main document","tool":"python","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":"Prepare file... | [{"item":"source_notes","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"main_document","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"file_index","check":"Check structure, completeness and consistency.","result":"simulated_chec... | {"status":"success","completed_outputs":["source_notes","main_document","file_index","email_draft"],"message":"The planned outputs are represented as completed in this synthetic example."} | document_workflow | train | true | 1.0.0 | ||
atm_0000005 | Plan a database migration with validation, rollback considerations and application compatibility. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Plan a database migration with validation, rollback considerations and application compatibility. Context: Sev... | ["terminal_operations","database_management","backend_engineering","software_testing","financial_analysis","project_management","document_creation"] | expert | recovered | ["file_manager","api_client","testing_framework","code_editor","database","email_client"] | ["Do not claim that an action succeeded without evidence.","Keep outputs structured and easy to inspect.","Validate calculations before presenting conclusions.","Explain uncertainty and missing information.","Report errors instead of silently ignoring them."] | ["migration_plan","schema_changes","validation_checks","rollback_plan"] | [{"step":1,"action":"Prepare migration plan","purpose":"Produce and validate migration plan","depends_on":[]},{"step":2,"action":"Prepare schema changes","purpose":"Produce and validate schema changes","depends_on":[1]},{"step":3,"action":"Prepare validation checks","purpose":"Produce and validate validation checks","d... | [{"step":1,"action":"Prepare migration plan","tool":"database","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare schema changes","tool":"testing_framework","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"actio... | {"code":"unavailable_service","description":"An external service cannot be confirmed as available."} | [{"item":"migration_plan","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"schema_changes","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"validation_checks","check":"Check structure, completeness and consistency.","result":"simu... | {"status":"recovered","completed_outputs":["migration_plan","schema_changes","validation_checks","rollback_plan"],"message":"The simulated issue was detected and a recovery path was completed.","recovery_actions":["Report the remaining limitation if recovery is not possible.","Use a documented alternative when authoriz... | database_migration | train | true | 1.0.0 | |
atm_0000006 | Collect source information, create structured documents and prepare organized deliverables. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Collect source information, create structured documents and prepare organized deliverables. Context: The outpu... | ["file_management","research","document_creation","email_writing"] | advanced | success | ["terminal","spreadsheet","file_manager","document_editor","testing_framework","database"] | ["Use accessible and understandable interfaces.","Keep outputs structured and easy to inspect.","Preserve original source data.","Do not claim that an action succeeded without evidence.","Explain uncertainty and missing information."] | ["source_notes","main_document","file_index","email_draft"] | [{"step":1,"action":"Prepare source notes","purpose":"Produce and validate source notes","depends_on":[]},{"step":2,"action":"Prepare main document","purpose":"Produce and validate main document","depends_on":[1]},{"step":3,"action":"Prepare file index","purpose":"Produce and validate file index","depends_on":[2]},{"st... | [{"step":1,"action":"Prepare source notes","tool":"terminal","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare main document","tool":"testing_framework","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":... | [{"item":"source_notes","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"main_document","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"file_index","check":"Check structure, completeness and consistency.","result":"simulated_chec... | {"status":"success","completed_outputs":["source_notes","main_document","file_index","email_draft"],"message":"The planned outputs are represented as completed in this synthetic example."} | document_workflow | train | true | 1.0.0 | ||
atm_0000007 | Research a topic, compare alternatives, prepare a structured proposal and draft an email. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Research a topic, compare alternatives, prepare a structured proposal and draft an email. Context: The project... | ["research","email_writing","document_creation","decision_making","software_testing","backend_engineering","data_analysis"] | very_complex | success | ["email_client","browser","code_editor","chart_generator","python"] | ["Do not claim that an action succeeded without evidence.","Respect dependencies between tasks.","Keep outputs structured and easy to inspect."] | ["research_summary","comparison","proposal","email_draft"] | [{"step":1,"action":"Prepare research summary","purpose":"Produce and validate research summary","depends_on":[]},{"step":2,"action":"Prepare comparison","purpose":"Produce and validate comparison","depends_on":[1]},{"step":3,"action":"Prepare proposal","purpose":"Produce and validate proposal","depends_on":[2]},{"step... | [{"step":1,"action":"Prepare research summary","tool":"code_editor","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare comparison","tool":"chart_generator","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action... | [{"item":"research_summary","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"comparison","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"proposal","check":"Check structure, completeness and consistency.","result":"simulated_check... | {"status":"success","completed_outputs":["research_summary","comparison","proposal","email_draft"],"message":"The planned outputs are represented as completed in this synthetic example."} | research_to_proposal | train | true | 1.0.0 | ||
atm_0000008 | Design and implement a user-friendly application with a backend, database and verification plan. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Design and implement a user-friendly application with a backend, database and verification plan. Context: The ... | ["app_development","backend_engineering","software_testing","database_management"] | expert | failed | ["file_manager","database","document_editor","api_client","code_editor","email_client","spreadsheet","testing_framework"] | ["Use accessible and understandable interfaces.","Validate calculations before presenting conclusions.","Preserve original source data."] | ["application_design","backend_plan","database_schema","test_report"] | [{"step":1,"action":"Prepare application design","purpose":"Produce and validate application design","depends_on":[]},{"step":2,"action":"Prepare backend plan","purpose":"Produce and validate backend plan","depends_on":[1]},{"step":3,"action":"Prepare database schema","purpose":"Produce and validate database schema","d... | [{"step":1,"action":"Prepare application design","tool":"spreadsheet","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare backend plan","tool":"testing_framework","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"... | {"code":"missing_input","description":"A required input is missing."} | [{"item":"application_design","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"backend_plan","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"database_schema","check":"Check structure, completeness and consistency.","result":"simu... | {"status":"failed","completed_outputs":[],"message":"The task could not be completed. The failure is reported explicitly.","next_step":"Investigate the error and obtain any missing authorization or input."} | fullstack_delivery | train | true | 1.0.0 | |
atm_0000009 | Plan a database migration with validation, rollback considerations and application compatibility. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Plan a database migration with validation, rollback considerations and application compatibility. Context: The... | ["terminal_operations","software_testing","database_management","backend_engineering"] | very_complex | failed | ["code_editor","browser","spreadsheet","python","document_editor","testing_framework","file_manager"] | ["Preserve original source data.","Use accessible and understandable interfaces.","Validate calculations before presenting conclusions.","Do not claim that an action succeeded without evidence.","Report errors instead of silently ignoring them."] | ["migration_plan","schema_changes","validation_checks","rollback_plan"] | [{"step":1,"action":"Prepare migration plan","purpose":"Produce and validate migration plan","depends_on":[]},{"step":2,"action":"Prepare schema changes","purpose":"Produce and validate schema changes","depends_on":[1]},{"step":3,"action":"Prepare validation checks","purpose":"Produce and validate validation checks","d... | [{"step":1,"action":"Prepare migration plan","tool":"document_editor","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare schema changes","tool":"code_editor","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"acti... | {"code":"invalid_format","description":"An input file has an unexpected format."} | [{"item":"migration_plan","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"schema_changes","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"validation_checks","check":"Check structure, completeness and consistency.","result":"simu... | {"status":"failed","completed_outputs":[],"message":"The task could not be completed. The failure is reported explicitly.","next_step":"Investigate the error and obtain any missing authorization or input."} | database_migration | train | true | 1.0.0 | |
atm_0000010 | Design and implement a user-friendly application with a backend, database and verification plan. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Design and implement a user-friendly application with a backend, database and verification plan. Context: Seve... | ["software_testing","database_management","app_development","backend_engineering","cross_domain_execution","email_writing","financial_analysis"] | intermediate | failed | ["terminal","testing_framework","spreadsheet","file_manager","python","code_editor","database"] | ["Respect dependencies between tasks.","Do not claim that an action succeeded without evidence.","Keep outputs structured and easy to inspect.","Avoid exposing credentials or private information.","Report errors instead of silently ignoring them."] | ["application_design","backend_plan","database_schema","test_report"] | [{"step":1,"action":"Prepare application design","purpose":"Produce and validate application design","depends_on":[]},{"step":2,"action":"Prepare backend plan","purpose":"Produce and validate backend plan","depends_on":[1]},{"step":3,"action":"Prepare database schema","purpose":"Produce and validate database schema","d... | [{"step":1,"action":"Prepare application design","tool":"python","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare backend plan","tool":"spreadsheet","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":"Pr... | {"code":"permission_denied","description":"The requested operation lacks authorization."} | [{"item":"application_design","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"backend_plan","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"database_schema","check":"Check structure, completeness and consistency.","result":"simu... | {"status":"failed","completed_outputs":[],"message":"The task could not be completed. The failure is reported explicitly.","next_step":"Investigate the error and obtain any missing authorization or input."} | fullstack_delivery | train | true | 1.0.0 | |
atm_0000011 | Interpret unfamiliar structured data, document limitations and prepare a reasoned recommendation. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Interpret unfamiliar structured data, document limitations and prepare a reasoned recommendation. Context: Sev... | ["financial_analysis","decision_making","data_analysis","document_creation","workflow_automation"] | very_complex | impossible | ["document_editor","testing_framework","file_manager","database","code_editor"] | ["Preserve original source data.","Report errors instead of silently ignoring them.","Validate calculations before presenting conclusions.","Respect dependencies between tasks.","Keep outputs structured and easy to inspect."] | ["data_assessment","calculations","limitations","recommendation"] | [{"step":1,"action":"Prepare data assessment","purpose":"Produce and validate data assessment","depends_on":[]},{"step":2,"action":"Prepare calculations","purpose":"Produce and validate calculations","depends_on":[1]},{"step":3,"action":"Prepare limitations","purpose":"Produce and validate limitations","depends_on":[2]... | [{"step":1,"action":"Prepare data assessment","tool":"file_manager","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare calculations","tool":"code_editor","status":"simulated","evidence":"Synthetic trace; no external action was executed."}] | [{"item":"data_assessment","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"calculations","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"limitations","check":"Check structure, completeness and consistency.","result":"simulated_c... | {"status":"impossible","completed_outputs":[],"reason":"The task requires information that has not been supplied.","message":"Explain the limitation and identify a feasible alternative."} | test_data_decision | test | true | 1.0.0 | ||
atm_0000012 | Evaluate information from multiple sources, identify uncertainty and create a concise report. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Evaluate information from multiple sources, identify uncertainty and create a concise report. Context: The wor... | ["data_analysis","research","document_creation"] | expert | recovered | ["email_client","document_editor","file_manager","code_editor"] | ["Use accessible and understandable interfaces.","Preserve original source data.","Explain uncertainty and missing information.","Respect dependencies between tasks.","Do not claim that an action succeeded without evidence."] | ["source_assessment","findings","uncertainties","report"] | [{"step":1,"action":"Prepare source assessment","purpose":"Produce and validate source assessment","depends_on":[]},{"step":2,"action":"Prepare findings","purpose":"Produce and validate findings","depends_on":[1]},{"step":3,"action":"Prepare uncertainties","purpose":"Produce and validate uncertainties","depends_on":[2]... | [{"step":1,"action":"Prepare source assessment","tool":"email_client","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare findings","tool":"code_editor","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":"P... | {"code":"test_failure","description":"A test did not meet its expected condition."} | [{"item":"source_assessment","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"findings","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"uncertainties","check":"Check structure, completeness and consistency.","result":"simulated_c... | {"status":"recovered","completed_outputs":["source_assessment","findings","uncertainties","report"],"message":"The simulated issue was detected and a recovery path was completed.","recovery_actions":["Preserve the original data before attempting a correction.","Request the missing input from the user.","Report the rema... | validation_research | validation | true | 1.0.0 | |
atm_0000013 | Analyse financial records, calculate key indicators, produce a report and prepare a communication. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Analyse financial records, calculate key indicators, produce a report and prepare a communication. Context: Th... | ["data_analysis","document_creation","email_writing","financial_analysis"] | expert | failed | ["api_client","spreadsheet","document_editor","email_client","python","file_manager"] | ["Respect dependencies between tasks.","Do not claim that an action succeeded without evidence.","Explain uncertainty and missing information.","Avoid exposing credentials or private information."] | ["validated_metrics","financial_report","executive_summary","email_draft"] | [{"step":1,"action":"Prepare validated metrics","purpose":"Produce and validate validated metrics","depends_on":[]},{"step":2,"action":"Prepare financial report","purpose":"Produce and validate financial report","depends_on":[1]},{"step":3,"action":"Prepare executive summary","purpose":"Produce and validate executive s... | [{"step":1,"action":"Prepare validated metrics","tool":"document_editor","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare financial report","tool":"email_client","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3... | {"code":"permission_denied","description":"The requested operation lacks authorization."} | [{"item":"validated_metrics","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"financial_report","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"executive_summary","check":"Check structure, completeness and consistency.","result":... | {"status":"failed","completed_outputs":[],"message":"The task could not be completed. The failure is reported explicitly.","next_step":"Investigate the error and obtain any missing authorization or input."} | financial_reporting | train | true | 1.0.0 | |
atm_0000014 | Inspect a software project, diagnose a reported problem, propose a correction and verify it. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Inspect a software project, diagnose a reported problem, propose a correction and verify it. Context: The avai... | ["file_management","terminal_operations","software_testing","debugging","cross_domain_execution"] | advanced | clarification | ["email_client","terminal","code_editor","database"] | ["Respect dependencies between tasks.","Keep outputs structured and easy to inspect.","Preserve original source data."] | ["diagnosis","repair_plan","test_results","change_summary"] | [{"step":1,"action":"Prepare diagnosis","purpose":"Produce and validate diagnosis","depends_on":[]},{"step":2,"action":"Prepare repair plan","purpose":"Produce and validate repair plan","depends_on":[1]},{"step":3,"action":"Prepare test results","purpose":"Produce and validate test results","depends_on":[2]},{"step":4,... | [{"step":1,"action":"Prepare diagnosis","tool":"email_client","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare repair plan","tool":"database","status":"simulated","evidence":"Synthetic trace; no external action was executed."}] | Should the application be designed for desktop, mobile or both? | [{"item":"diagnosis","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"repair_plan","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"test_results","check":"Check structure, completeness and consistency.","result":"simulated_check"}... | {"status":"needs_clarification","completed_outputs":[],"question":"Should the application be designed for desktop, mobile or both?","message":"Pause consequential execution until the ambiguity is resolved."} | terminal_debugging | train | true | 1.0.0 | |
atm_0000015 | Coordinate a project with deadlines, dependencies, stakeholders and documented decisions. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Coordinate a project with deadlines, dependencies, stakeholders and documented decisions. Context: The user ex... | ["email_writing","project_management","document_creation","decision_making","debugging"] | advanced | recovered | ["terminal","email_client","chart_generator","browser","spreadsheet"] | ["Report errors instead of silently ignoring them.","Validate calculations before presenting conclusions.","Explain uncertainty and missing information."] | ["project_plan","timeline","decision_log","communications"] | [{"step":1,"action":"Prepare project plan","purpose":"Produce and validate project plan","depends_on":[]},{"step":2,"action":"Prepare timeline","purpose":"Produce and validate timeline","depends_on":[1]},{"step":3,"action":"Prepare decision log","purpose":"Produce and validate decision log","depends_on":[2]},{"step":4,... | [{"step":1,"action":"Prepare project plan","tool":"terminal","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare timeline","tool":"browser","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":"Prepare decisi... | {"code":"dependency_error","description":"A required dependency is unavailable."} | [{"item":"project_plan","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"timeline","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"decision_log","check":"Check structure, completeness and consistency.","result":"simulated_check"}... | {"status":"recovered","completed_outputs":["project_plan","timeline","decision_log","communications"],"message":"The simulated issue was detected and a recovery path was completed.","recovery_actions":["Request the missing input from the user.","Report the remaining limitation if recovery is not possible."]} | project_coordination | train | true | 1.0.0 | |
atm_0000016 | Investigate a disrupted workflow, isolate the problem and prepare a verified recovery plan. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Investigate a disrupted workflow, isolate the problem and prepare a verified recovery plan. Context: Some assu... | ["debugging","terminal_operations","project_management","software_testing"] | very_complex | failed | ["file_manager","document_editor","python","spreadsheet","database","email_client","terminal","chart_generator"] | ["Preserve original source data.","Respect dependencies between tasks.","Validate calculations before presenting conclusions."] | ["incident_summary","root_cause_hypothesis","recovery_plan","verification_report"] | [{"step":1,"action":"Prepare incident summary","purpose":"Produce and validate incident summary","depends_on":[]},{"step":2,"action":"Prepare root cause hypothesis","purpose":"Produce and validate root cause hypothesis","depends_on":[1]},{"step":3,"action":"Prepare recovery plan","purpose":"Produce and validate recover... | [{"step":1,"action":"Prepare incident summary","tool":"chart_generator","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare root cause hypothesis","tool":"database","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3... | {"code":"ambiguous_requirement","description":"The requirement has multiple plausible meanings."} | [{"item":"incident_summary","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"root_cause_hypothesis","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"recovery_plan","check":"Check structure, completeness and consistency.","result":... | {"status":"failed","completed_outputs":[],"message":"The task could not be completed. The failure is reported explicitly.","next_step":"Investigate the error and obtain any missing authorization or input."} | test_recovery_workflow | test | true | 1.0.0 | |
atm_0000017 | Research a topic, compare alternatives, prepare a structured proposal and draft an email. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Research a topic, compare alternatives, prepare a structured proposal and draft an email. Context: The task re... | ["decision_making","email_writing","research","document_creation"] | expert | recovered | ["terminal","email_client","file_manager","document_editor","chart_generator"] | ["Keep outputs structured and easy to inspect.","Avoid exposing credentials or private information.","Explain uncertainty and missing information.","Validate calculations before presenting conclusions."] | ["research_summary","comparison","proposal","email_draft"] | [{"step":1,"action":"Prepare research summary","purpose":"Produce and validate research summary","depends_on":[]},{"step":2,"action":"Prepare comparison","purpose":"Produce and validate comparison","depends_on":[1]},{"step":3,"action":"Prepare proposal","purpose":"Produce and validate proposal","depends_on":[2]},{"step... | [{"step":1,"action":"Prepare research summary","tool":"terminal","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare comparison","tool":"terminal","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":"Prepare... | {"code":"missing_input","description":"A required input is missing."} | [{"item":"research_summary","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"comparison","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"proposal","check":"Check structure, completeness and consistency.","result":"simulated_check... | {"status":"recovered","completed_outputs":["research_summary","comparison","proposal","email_draft"],"message":"The simulated issue was detected and a recovery path was completed.","recovery_actions":["Report the remaining limitation if recovery is not possible.","Correct the issue and repeat the relevant validation."]... | research_to_proposal | train | true | 1.0.0 | |
atm_0000018 | Evaluate information from multiple sources, identify uncertainty and create a concise report. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Evaluate information from multiple sources, identify uncertainty and create a concise report. Context: The out... | ["data_analysis","document_creation","research"] | intermediate | success | ["chart_generator","document_editor","email_client","api_client","database","code_editor","testing_framework","python"] | ["Validate calculations before presenting conclusions.","Preserve original source data.","Avoid exposing credentials or private information."] | ["source_assessment","findings","uncertainties","report"] | [{"step":1,"action":"Prepare source assessment","purpose":"Produce and validate source assessment","depends_on":[]},{"step":2,"action":"Prepare findings","purpose":"Produce and validate findings","depends_on":[1]},{"step":3,"action":"Prepare uncertainties","purpose":"Produce and validate uncertainties","depends_on":[2]... | [{"step":1,"action":"Prepare source assessment","tool":"email_client","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare findings","tool":"api_client","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":"Pr... | [{"item":"source_assessment","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"findings","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"uncertainties","check":"Check structure, completeness and consistency.","result":"simulated_c... | {"status":"success","completed_outputs":["source_assessment","findings","uncertainties","report"],"message":"The planned outputs are represented as completed in this synthetic example."} | validation_research | validation | true | 1.0.0 | ||
atm_0000019 | Plan a small software product from requirements through implementation, testing and documentation. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Plan a small software product from requirements through implementation, testing and documentation. Context: Th... | ["software_testing","app_development","project_management","document_creation","backend_engineering","financial_analysis"] | advanced | success | ["chart_generator","email_client","database","document_editor","testing_framework","python"] | ["Explain uncertainty and missing information.","Avoid exposing credentials or private information.","Respect dependencies between tasks.","Keep outputs structured and easy to inspect.","Validate calculations before presenting conclusions."] | ["requirements","architecture","implementation_plan","test_report","documentation"] | [{"step":1,"action":"Prepare requirements","purpose":"Produce and validate requirements","depends_on":[]},{"step":2,"action":"Prepare architecture","purpose":"Produce and validate architecture","depends_on":[1]},{"step":3,"action":"Prepare implementation plan","purpose":"Produce and validate implementation plan","depen... | [{"step":1,"action":"Prepare requirements","tool":"testing_framework","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare architecture","tool":"database","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":"... | [{"item":"requirements","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"architecture","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"implementation_plan","check":"Check structure, completeness and consistency.","result":"simula... | {"status":"success","completed_outputs":["requirements","architecture","implementation_plan","test_report","documentation"],"message":"The planned outputs are represented as completed in this synthetic example."} | integrated_product | train | true | 1.0.0 | ||
atm_0000020 | Analyse financial records, calculate key indicators, produce a report and prepare a communication. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Analyse financial records, calculate key indicators, produce a report and prepare a communication. Context: Th... | ["email_writing","data_analysis","document_creation","financial_analysis","workflow_automation"] | intermediate | recovered | ["document_editor","browser","database","spreadsheet","testing_framework","api_client","file_manager","email_client"] | ["Do not claim that an action succeeded without evidence.","Respect dependencies between tasks.","Use accessible and understandable interfaces."] | ["validated_metrics","financial_report","executive_summary","email_draft"] | [{"step":1,"action":"Prepare validated metrics","purpose":"Produce and validate validated metrics","depends_on":[]},{"step":2,"action":"Prepare financial report","purpose":"Produce and validate financial report","depends_on":[1]},{"step":3,"action":"Prepare executive summary","purpose":"Produce and validate executive s... | [{"step":1,"action":"Prepare validated metrics","tool":"file_manager","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare financial report","tool":"browser","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action... | {"code":"test_failure","description":"A test did not meet its expected condition."} | [{"item":"validated_metrics","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"financial_report","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"executive_summary","check":"Check structure, completeness and consistency.","result":... | {"status":"recovered","completed_outputs":["validated_metrics","financial_report","executive_summary","email_draft"],"message":"The simulated issue was detected and a recovery path was completed.","recovery_actions":["Preserve the original data before attempting a correction.","Request the missing input from the user."... | financial_reporting | train | true | 1.0.0 | |
atm_0000021 | Design an API-backed product with clear interfaces, input validation and tests. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Design an API-backed product with clear interfaces, input validation and tests. Context: The available informa... | ["backend_engineering","app_development","api_integration","software_testing","file_management","document_creation","ui_ux_design"] | expert | clarification | ["file_manager","spreadsheet","database","document_editor","testing_framework","code_editor","browser"] | ["Keep outputs structured and easy to inspect.","Preserve original source data.","Report errors instead of silently ignoring them.","Avoid exposing credentials or private information."] | ["api_contract","backend_design","interface_spec","test_plan"] | [{"step":1,"action":"Prepare api contract","purpose":"Produce and validate api contract","depends_on":[]},{"step":2,"action":"Prepare backend design","purpose":"Produce and validate backend design","depends_on":[1]},{"step":3,"action":"Prepare interface spec","purpose":"Produce and validate interface spec","depends_on"... | [] | Which source should be treated as authoritative? | [{"item":"api_contract","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"backend_design","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"interface_spec","check":"Check structure, completeness and consistency.","result":"simulated... | {"status":"needs_clarification","completed_outputs":[],"question":"Which source should be treated as authoritative?","message":"Pause consequential execution until the ambiguity is resolved."} | api_product | train | true | 1.0.0 | |
atm_0000022 | Collect source information, create structured documents and prepare organized deliverables. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Collect source information, create structured documents and prepare organized deliverables. Context: The avail... | ["document_creation","research","file_management","email_writing"] | expert | failed | ["spreadsheet","document_editor","email_client","database","browser"] | ["Report errors instead of silently ignoring them.","Preserve original source data.","Avoid exposing credentials or private information."] | ["source_notes","main_document","file_index","email_draft"] | [{"step":1,"action":"Prepare source notes","purpose":"Produce and validate source notes","depends_on":[]},{"step":2,"action":"Prepare main document","purpose":"Produce and validate main document","depends_on":[1]},{"step":3,"action":"Prepare file index","purpose":"Produce and validate file index","depends_on":[2]},{"st... | [{"step":1,"action":"Prepare source notes","tool":"browser","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare main document","tool":"database","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":"Prepare f... | {"code":"dependency_error","description":"A required dependency is unavailable."} | [{"item":"source_notes","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"main_document","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"file_index","check":"Check structure, completeness and consistency.","result":"simulated_chec... | {"status":"failed","completed_outputs":[],"message":"The task could not be completed. The failure is reported explicitly.","next_step":"Investigate the error and obtain any missing authorization or input."} | document_workflow | train | true | 1.0.0 | |
atm_0000023 | Inspect a software project, diagnose a reported problem, propose a correction and verify it. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Inspect a software project, diagnose a reported problem, propose a correction and verify it. Context: The avai... | ["file_management","debugging","software_testing","terminal_operations"] | very_complex | failed | ["spreadsheet","api_client","document_editor","file_manager","chart_generator"] | ["Preserve original source data.","Avoid exposing credentials or private information.","Validate calculations before presenting conclusions.","Explain uncertainty and missing information.","Do not claim that an action succeeded without evidence."] | ["diagnosis","repair_plan","test_results","change_summary"] | [{"step":1,"action":"Prepare diagnosis","purpose":"Produce and validate diagnosis","depends_on":[]},{"step":2,"action":"Prepare repair plan","purpose":"Produce and validate repair plan","depends_on":[1]},{"step":3,"action":"Prepare test results","purpose":"Produce and validate test results","depends_on":[2]},{"step":4,... | [{"step":1,"action":"Prepare diagnosis","tool":"document_editor","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare repair plan","tool":"chart_generator","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":... | {"code":"test_failure","description":"A test did not meet its expected condition."} | [{"item":"diagnosis","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"repair_plan","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"test_results","check":"Check structure, completeness and consistency.","result":"simulated_check"}... | {"status":"failed","completed_outputs":[],"message":"The task could not be completed. The failure is reported explicitly.","next_step":"Investigate the error and obtain any missing authorization or input."} | terminal_debugging | train | true | 1.0.0 | |
atm_0000024 | Investigate a disrupted workflow, isolate the problem and prepare a verified recovery plan. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Investigate a disrupted workflow, isolate the problem and prepare a verified recovery plan. Context: The workf... | ["project_management","debugging","software_testing","terminal_operations"] | expert | success | ["email_client","database","document_editor","python","browser","spreadsheet","api_client","terminal"] | ["Explain uncertainty and missing information.","Use accessible and understandable interfaces.","Preserve original source data.","Avoid exposing credentials or private information.","Validate calculations before presenting conclusions."] | ["incident_summary","root_cause_hypothesis","recovery_plan","verification_report"] | [{"step":1,"action":"Prepare incident summary","purpose":"Produce and validate incident summary","depends_on":[]},{"step":2,"action":"Prepare root cause hypothesis","purpose":"Produce and validate root cause hypothesis","depends_on":[1]},{"step":3,"action":"Prepare recovery plan","purpose":"Produce and validate recover... | [{"step":1,"action":"Prepare incident summary","tool":"document_editor","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare root cause hypothesis","tool":"api_client","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step"... | [{"item":"incident_summary","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"root_cause_hypothesis","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"recovery_plan","check":"Check structure, completeness and consistency.","result":... | {"status":"success","completed_outputs":["incident_summary","root_cause_hypothesis","recovery_plan","verification_report"],"message":"The planned outputs are represented as completed in this synthetic example."} | test_recovery_workflow | test | true | 1.0.0 | ||
atm_0000025 | Design an automated workflow with dependencies, error handling and organized outputs. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Design an automated workflow with dependencies, error handling and organized outputs. Context: The task requir... | ["project_management","file_management","workflow_automation","api_integration","terminal_operations"] | very_complex | clarification | ["document_editor","email_client","spreadsheet","testing_framework","database","terminal","file_manager","api_client"] | ["Avoid exposing credentials or private information.","Keep outputs structured and easy to inspect.","Respect dependencies between tasks."] | ["workflow_specification","dependency_map","error_policy","implementation_plan"] | [{"step":1,"action":"Prepare workflow specification","purpose":"Produce and validate workflow specification","depends_on":[]},{"step":2,"action":"Prepare dependency map","purpose":"Produce and validate dependency map","depends_on":[1]},{"step":3,"action":"Prepare error policy","purpose":"Produce and validate error poli... | [{"step":1,"action":"Prepare workflow specification","tool":"spreadsheet","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare dependency map","tool":"spreadsheet","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"... | Which output format should be used? | [{"item":"workflow_specification","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"dependency_map","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"error_policy","check":"Check structure, completeness and consistency.","result":"s... | {"status":"needs_clarification","completed_outputs":[],"question":"Which output format should be used?","message":"Pause consequential execution until the ambiguity is resolved."} | business_automation | train | true | 1.0.0 | |
atm_0000026 | Research a topic, compare alternatives, prepare a structured proposal and draft an email. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Research a topic, compare alternatives, prepare a structured proposal and draft an email. Context: The user ex... | ["email_writing","research","decision_making","document_creation","debugging"] | expert | clarification | ["chart_generator","spreadsheet","python","document_editor","api_client","file_manager","terminal"] | ["Do not claim that an action succeeded without evidence.","Use accessible and understandable interfaces.","Avoid exposing credentials or private information."] | ["research_summary","comparison","proposal","email_draft"] | [{"step":1,"action":"Prepare research summary","purpose":"Produce and validate research summary","depends_on":[]},{"step":2,"action":"Prepare comparison","purpose":"Produce and validate comparison","depends_on":[1]},{"step":3,"action":"Prepare proposal","purpose":"Produce and validate proposal","depends_on":[2]},{"step... | [{"step":1,"action":"Prepare research summary","tool":"api_client","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare comparison","tool":"chart_generator","status":"simulated","evidence":"Synthetic trace; no external action was executed."}] | Can you confirm the intended recipient and tone? | [{"item":"research_summary","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"comparison","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"proposal","check":"Check structure, completeness and consistency.","result":"simulated_check... | {"status":"needs_clarification","completed_outputs":[],"question":"Can you confirm the intended recipient and tone?","message":"Pause consequential execution until the ambiguity is resolved."} | research_to_proposal | train | true | 1.0.0 | |
atm_0000027 | Inspect a software project, diagnose a reported problem, propose a correction and verify it. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Inspect a software project, diagnose a reported problem, propose a correction and verify it. Context: The user... | ["terminal_operations","debugging","file_management","software_testing","document_creation"] | advanced | success | ["file_manager","terminal","database","code_editor","chart_generator","python","browser"] | ["Report errors instead of silently ignoring them.","Respect dependencies between tasks.","Avoid exposing credentials or private information.","Preserve original source data."] | ["diagnosis","repair_plan","test_results","change_summary"] | [{"step":1,"action":"Prepare diagnosis","purpose":"Produce and validate diagnosis","depends_on":[]},{"step":2,"action":"Prepare repair plan","purpose":"Produce and validate repair plan","depends_on":[1]},{"step":3,"action":"Prepare test results","purpose":"Produce and validate test results","depends_on":[2]},{"step":4,... | [{"step":1,"action":"Prepare diagnosis","tool":"browser","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare repair plan","tool":"file_manager","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":"Prepare te... | [{"item":"diagnosis","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"repair_plan","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"test_results","check":"Check structure, completeness and consistency.","result":"simulated_check"}... | {"status":"success","completed_outputs":["diagnosis","repair_plan","test_results","change_summary"],"message":"The planned outputs are represented as completed in this synthetic example."} | terminal_debugging | train | true | 1.0.0 | ||
atm_0000028 | Plan a database migration with validation, rollback considerations and application compatibility. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Plan a database migration with validation, rollback considerations and application compatibility. Context: The... | ["backend_engineering","database_management","terminal_operations","software_testing"] | advanced | recovered | ["api_client","database","terminal","file_manager"] | ["Keep outputs structured and easy to inspect.","Validate calculations before presenting conclusions.","Explain uncertainty and missing information.","Do not claim that an action succeeded without evidence.","Respect dependencies between tasks."] | ["migration_plan","schema_changes","validation_checks","rollback_plan"] | [{"step":1,"action":"Prepare migration plan","purpose":"Produce and validate migration plan","depends_on":[]},{"step":2,"action":"Prepare schema changes","purpose":"Produce and validate schema changes","depends_on":[1]},{"step":3,"action":"Prepare validation checks","purpose":"Produce and validate validation checks","d... | [{"step":1,"action":"Prepare migration plan","tool":"api_client","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare schema changes","tool":"file_manager","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action":... | {"code":"missing_input","description":"A required input is missing."} | [{"item":"migration_plan","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"schema_changes","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"validation_checks","check":"Check structure, completeness and consistency.","result":"simu... | {"status":"recovered","completed_outputs":["migration_plan","schema_changes","validation_checks","rollback_plan"],"message":"The simulated issue was detected and a recovery path was completed.","recovery_actions":["Request the missing input from the user.","Preserve the original data before attempting a correction.","I... | database_migration | train | true | 1.0.0 | |
atm_0000029 | Coordinate a complex application delivery with technical dependencies and stakeholder communication. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Coordinate a complex application delivery with technical dependencies and stakeholder communication. Context: ... | ["app_development","project_management","email_writing","backend_engineering","database_management","debugging"] | advanced | recovered | ["spreadsheet","email_client","code_editor","terminal"] | ["Validate calculations before presenting conclusions.","Report errors instead of silently ignoring them.","Use accessible and understandable interfaces.","Respect dependencies between tasks.","Preserve original source data."] | ["delivery_plan","technical_design","dependency_report","email_draft"] | [{"step":1,"action":"Prepare delivery plan","purpose":"Produce and validate delivery plan","depends_on":[]},{"step":2,"action":"Prepare technical design","purpose":"Produce and validate technical design","depends_on":[1]},{"step":3,"action":"Prepare dependency report","purpose":"Produce and validate dependency report",... | [{"step":1,"action":"Prepare delivery plan","tool":"code_editor","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":2,"action":"Prepare technical design","tool":"spreadsheet","status":"simulated","evidence":"Synthetic trace; no external action was executed."},{"step":3,"action"... | {"code":"invalid_format","description":"An input file has an unexpected format."} | [{"item":"delivery_plan","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"technical_design","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"dependency_report","check":"Check structure, completeness and consistency.","result":"sim... | {"status":"recovered","completed_outputs":["delivery_plan","technical_design","dependency_report","email_draft"],"message":"The simulated issue was detected and a recovery path was completed.","recovery_actions":["Correct the issue and repeat the relevant validation.","Inspect the failing component and isolate the caus... | test_complex_delivery | test | true | 1.0.0 | |
atm_0000030 | Design an automated workflow with dependencies, error handling and organized outputs. | [{"role":"system","content":"You are a task-execution agent. Plan carefully, respect constraints, verify outputs, report uncertainty, and never claim an unexecuted action succeeded."},{"role":"user","content":"Design an automated workflow with dependencies, error handling and organized outputs. Context: Some assumption... | ["workflow_automation","file_management","project_management","api_integration"] | intermediate | impossible | ["browser","document_editor","code_editor","email_client","testing_framework"] | ["Report errors instead of silently ignoring them.","Do not claim that an action succeeded without evidence.","Preserve original source data.","Validate calculations before presenting conclusions."] | ["workflow_specification","dependency_map","error_policy","implementation_plan"] | [{"step":1,"action":"Prepare workflow specification","purpose":"Produce and validate workflow specification","depends_on":[]},{"step":2,"action":"Prepare dependency map","purpose":"Produce and validate dependency map","depends_on":[1]},{"step":3,"action":"Prepare error policy","purpose":"Produce and validate error poli... | [{"step":1,"action":"Prepare workflow specification","tool":"testing_framework","status":"simulated","evidence":"Synthetic trace; no external action was executed."}] | [{"item":"workflow_specification","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"dependency_map","check":"Check structure, completeness and consistency.","result":"simulated_check"},{"item":"error_policy","check":"Check structure, completeness and consistency.","result":"s... | {"status":"impossible","completed_outputs":[],"reason":"The requested outputs contradict a stated constraint.","message":"Explain the limitation and identify a feasible alternative."} | business_automation | train | true | 1.0.0 |
End of preview. Expand in Data Studio
Agent Task Management Dataset
5,000,000 compact synthetic examples for agent task execution.
Repository: Lelonthecodeur/agent-task-management-dataset
Purpose
This dataset is designed to train general-purpose agents to plan, coordinate, execute, verify and report complex tasks involving multiple tools and domains.
Domains
- email_writing
- document_creation
- app_development
- terminal_operations
- backend_engineering
- financial_analysis
- data_analysis
- project_management
- research
- workflow_automation
- software_testing
- ui_ux_design
- file_management
- decision_making
- database_management
- api_integration
- debugging
- cross_domain_execution
Scenario types
The dataset includes:
- Successful task completion.
- Failed attempts.
- Error detection and recovery.
- Impossible or unsupported tasks.
- Clarification and ambiguity handling.
Example structure
Each Parquet record contains:
idtaskmessagesdomainsdifficultyoutcomeavailable_toolsconstraintsexpected_outputsexecution_plantool_traceerrorclarificationverificationfinal_resultscenario_familysplitsyntheticgenerator_version
The messages field contains a compact JSON conversation
with system, user and assistant messages.
Dataset splits
The dataset uses approximately:
- 80% training
- 10% validation
- 10% test
Splits are sampled independently. Scenario families are not strictly held out across splits, so this dataset should not be treated as a leakage-proof benchmark.
Synthetic-data disclosure
All examples are programmatically generated.
Tool traces are simulated representations. No real terminal commands, emails, financial transactions, API requests, database changes or deployments were performed.
The examples do not establish that a task was actually completed in an external environment.
The generator uses structured templates and randomized combinations. Consequently, some examples may be repetitive, simplified or less realistic than expert-written trajectories.
Difficulty labels are assigned by the generator and are not independently benchmarked.
This dataset is a synthetic training resource, not a substitute for verified real-world agent execution data.
Intended use
- Agent planning and task decomposition.
- Multi-domain workflow training.
- Tool-selection reasoning.
- Error handling and recovery.
- Clarification behavior.
- Output verification.
- Structured task reporting.
Limitations
Do not interpret simulated execution traces as real execution evidence. Verify model behavior independently before deploying systems trained on this dataset.
Generation details
- Examples: 5,000,000
- Format: Parquet
- Compression: Zstandard
- Generator version: 1.0.0
- License: MIT
License
MIT License. See the LICENSE file.
- Downloads last month
- 42