ssaraf1/slm-workflow-planner-7b-v1
Text Generation • Updated
messages listlengths 3 3 |
|---|
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type JOIN, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: JOIN\n\nCandidates:\n 1. INITIAL_TRIAGE (AGENT) [processor] → produces: severity_level, claim_category, estimated_amount, capabilities: produces_severity,produces_claim,produces_estimated\n\nState:\n goal_progress=0.23\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.45\n uncertainty_level=0.00\n cost_accrued=501.50\n\nSelect join node:"
},
{
"role": "assistant",
"content": "INITIAL_TRIAGE"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type JOIN, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: JOIN\n\nCandidates:\n 1. FRAUD_SCREENING (SYSTEM) [processor] → produces: fraud_score, fraud_flags, capabilities: produces_fraud,can_succeed,outcome_screening_clear\n\nState:\n goal_progress=0.32\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.30\n uncertainty_level=0.90\n cost_accrued=46.00\n\nSelect join node:"
},
{
"role": "assistant",
"content": "FRAUD_SCREENING"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: BRANCH_3 (HUMAN)\nOutcome: failure\n\nState:\n goal_progress=0.40\n retry_count=1\n total_retries=1\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=2\n nodes_visited=2\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=62.00\n cost_accrued=50.50\n uncertainty_level=0.55\n\nEligible nodes:\n 1. BRANCH_3 (HUMAN) [retry_handler] → produces: ctx_b3 intent:human_task\n 2. JOIN (SYSTEM) [processor] → produces: ctx_joined intent:automated\n 3. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_FAIL, JOIN}]\nJoin-ready: ['BRANCH_3']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "FORK"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_2 (AGENT)\nOutcome: retry\n\nState:\n goal_progress=0.67\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=1\n nodes_visited=2\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=17.00\n cost_accrued=10.50\n uncertainty_level=0.20\n cost_budget=4990\n\nEligible nodes:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step_3 intent:human_task\n 2. STEP_2 (AGENT) [retry_handler] → produces: ctx_step_2\n 3. END_ESCALATED (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n\nForkable sets: [{END_ESCALATED, STEP_3}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "RETRY"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: BRANCH_2 (AGENT)\nOutcome: retry\n\nState:\n goal_progress=0.50\n retry_count=1\n total_retries=1\n parallel_active=2\n parallel_depth=1\n parallel_completed=0\n steps_taken=3\n nodes_visited=4\n escalation_count=0\n resource_pressure=0.30\n sla_pressure=0.00\n elapsed_time_minutes=47.00\n cost_accrued=30.50\n uncertainty_level=0.40\n\nEligible nodes:\n 1. BRANCH_2 (AGENT) [retry_handler] → produces: ctx_b2\n 2. JOIN (SYSTEM) [processor] → produces: ctx_joined intent:automated\n 3. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_FAIL, JOIN}]\nJoin-ready: ['BRANCH_2']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "JOIN"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type RETRY, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: RETRY\n\nCandidates:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step_3, capabilities: produces_ctx,can_succeed,outcome_success\n 2. STEP_2 (AGENT) [retry_handler] → produces: ctx_step_2, capabilities: produces_ctx,can_succeed,outcome_success\n 3. END_ESCALATED (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=0.40\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.30\n uncertainty_level=0.15\n cost_accrued=30.50\n\nSelect node for RETRY:"
},
{
"role": "assistant",
"content": "STEP_2"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: VERIFY_COMPLIANCE_V19 (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=0.29\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=1\n nodes_visited=2\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=4.00\n cost_accrued=1.00\n uncertainty_level=0.10\n\nEligible nodes:\n 1. MERGE_RESULTS (SYSTEM) [processor] → produces: ctx_merged intent:automated\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type JOIN, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: JOIN\n\nCandidates:\n 1. INITIAL_TRIAGE (AGENT) [processor] → produces: severity_level, claim_category, estimated_amount, capabilities: produces_severity,produces_claim,produces_estimated\n\nState:\n goal_progress=0.23\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.45\n uncertainty_level=0.00\n cost_accrued=120.50\n\nSelect join node:"
},
{
"role": "assistant",
"content": "INITIAL_TRIAGE"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step_3, capabilities: produces_ctx,can_succeed,outcome_success\n 2. STEP_2 (AGENT) [retry_handler] → produces: ctx_step_2, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.29\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.10\n cost_accrued=10.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "STEP_3"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. BRANCH_1 (SYSTEM) [processor] → produces: ctx_b1, capabilities: produces_ctx,can_succeed,outcome_success\n 2. BRANCH_2 (AGENT) [processor] → produces: ctx_b2, capabilities: produces_ctx,can_succeed,outcome_success\n 3. BRANCH_3 (HUMAN) [processor] → produces: ctx_b3, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.20\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=0.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "BRANCH_3"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: BRANCH_3 (HUMAN)\nOutcome: retry\n\nState:\n goal_progress=0.40\n retry_count=0\n total_retries=0\n parallel_active=2\n parallel_depth=0\n parallel_completed=0\n steps_taken=14\n nodes_visited=4\n escalation_count=0\n resource_pressure=0.30\n sla_pressure=0.00\n elapsed_time_minutes=422.00\n cost_accrued=350.50\n uncertainty_level=0.00\n\nEligible nodes:\n 1. BRANCH_3 (HUMAN) [retry_handler] → produces: ctx_b3 intent:human_task\n 2. JOIN (SYSTEM) [processor] → produces: ctx_joined intent:automated\n 3. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_FAIL, JOIN}]\nJoin-ready: ['BRANCH_3']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "JOIN"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type RETRY, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: RETRY\n\nCandidates:\n 1. STAGE_2 (AGENT) [retry_handler] → produces: ctx_stage2, capabilities: produces_ctx,can_succeed,outcome_success\n 2. STAGE_3 (HUMAN) [processor] → produces: ctx_stage3, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.20\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.30\n cost_accrued=10.50\n\nSelect node for RETRY:"
},
{
"role": "assistant",
"content": "STAGE_2"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type FORK, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: FORK\n\nCandidates:\n 1. APPROVE_WORK_V14 (AGENT) [gatekeeper] → produces: approval_status, capabilities: produces_approval,can_succeed,outcome_success\n 2. END_FAIL (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n 3. END_ESCALATED_INSPECT_CONDITION_V14 (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=0.40\n parallel_depth=0\n retry_count=0\n sla_pressure=0.13\n resource_pressure=0.00\n uncertainty_level=0.20\n cost_accrued=25.50\n\nSelect nodes to fork:"
},
{
"role": "assistant",
"content": "APPROVE_WORK_V14, END_ESCALATED_INSPECT_CONDITION_V14"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_5 (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=1.00\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=0\n nodes_visited=0\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=0.00\n cost_accrued=0.00\n uncertainty_level=0.00\n\nEligible nodes:\n 1. END_SUCCESS (SYSTEM) → produces: none\n 2. END_FAIL (SYSTEM) → produces: none\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "META"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. STEP_3 (HUMAN) [retry_handler] → produces: ctx_step3, capabilities: produces_ctx,can_succeed,outcome_success\n 2. STEP_4 (SYSTEM) [processor] → produces: ctx_step4, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.75\n parallel_depth=0\n retry_count=1\n sla_pressure=0.32\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=60.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "STEP_4"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: JOIN (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=0.60\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=2\n nodes_visited=3\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.14\n elapsed_time_minutes=34.00\n cost_accrued=26.00\n uncertainty_level=0.15\n time_to_deadline_minutes=206\n\nEligible nodes:\n 1. END_SUCCESS (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_2 (AGENT)\nOutcome: retry\n\nState:\n goal_progress=0.40\n retry_count=0\n total_retries=0\n parallel_active=2\n parallel_depth=0\n parallel_completed=0\n steps_taken=7\n nodes_visited=4\n escalation_count=0\n resource_pressure=0.30\n sla_pressure=0.00\n elapsed_time_minutes=107.00\n cost_accrued=70.50\n uncertainty_level=0.15\n\nEligible nodes:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step_3 intent:human_task\n 2. STEP_2 (AGENT) [retry_handler] → produces: ctx_step_2\n 3. END_ESCALATED (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n\nForkable sets: [{END_ESCALATED, STEP_3}]\nJoin-ready: ['STEP_2']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "JOIN"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step_3, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.67\n parallel_depth=0\n retry_count=1\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.10\n cost_accrued=20.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "STEP_3"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type FORK, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: FORK\n\nCandidates:\n 1. STEP_5 (AGENT) [processor] → produces: ctx_step_5, capabilities: produces_ctx,can_succeed,outcome_success\n 2. END_EXIT_4 (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=0.57\n parallel_depth=0\n retry_count=0\n sla_pressure=0.03\n resource_pressure=0.00\n uncertainty_level=0.10\n cost_accrued=36.00\n\nSelect nodes to fork:"
},
{
"role": "assistant",
"content": "END_EXIT_4, STEP_5"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type JOIN, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: JOIN\n\nCandidates:\n 1. INITIAL_TRIAGE (AGENT) [processor] → produces: severity_level, claim_category, estimated_amount, capabilities: produces_severity,produces_claim,produces_estimated\n\nState:\n goal_progress=0.23\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.45\n uncertainty_level=0.00\n cost_accrued=890.50\n\nSelect join node:"
},
{
"role": "assistant",
"content": "INITIAL_TRIAGE"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: JOIN (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=0.75\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=2\n nodes_visited=3\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=6.00\n cost_accrued=1.50\n uncertainty_level=0.05\n\nEligible nodes:\n 1. END_SUCCESS (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: DIAGNOSTIC_SCAN_V14 (DEVICE)\nOutcome: retry\n\nState:\n goal_progress=0.33\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=1\n nodes_visited=2\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=12.00\n cost_accrued=5.50\n uncertainty_level=0.40\n\nEligible nodes:\n 1. DIAGNOSTIC_SCAN_V14 (DEVICE) [retry_handler] → produces: scan_result\n 2. MERGE_RESULTS (SYSTEM) [processor] → produces: ctx_merged intent:automated\n 3. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "RETRY"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STAGE_2 (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=0.20\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=0\n nodes_visited=0\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=0.00\n cost_accrued=0.00\n uncertainty_level=0.00\n\nEligible nodes:\n 1. STAGE_2 (AGENT) → produces: none\n 2. STAGE_3 (HUMAN) → produces: none\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "META"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type JOIN, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: JOIN\n\nCandidates:\n 1. STEP_2 (AGENT) [retry_handler] → produces: ctx_step_2, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.40\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.30\n uncertainty_level=0.00\n cost_accrued=170.50\n\nSelect join node:"
},
{
"role": "assistant",
"content": "STEP_2"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_2 (AGENT)\nOutcome: failure\n\nState:\n goal_progress=0.67\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=1\n nodes_visited=2\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.07\n elapsed_time_minutes=17.00\n cost_accrued=10.50\n uncertainty_level=0.35\n time_to_deadline_minutes=223\n\nEligible nodes:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step_3 intent:human_task\n 2. END_EXIT_2 (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_EXIT_2, STEP_3}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "FORK"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. NOTIFY_CUSTOMER_V11 (SYSTEM) [communicator] → produces: notif_id, capabilities: produces_notif,can_succeed,outcome_success\n 2. LOG_AUDIT_TRAIL_V11 (SYSTEM) [processor] → produces: audit_id, capabilities: produces_audit,can_succeed,outcome_success\n 3. SEND_CONFIRMATION_V11 (SYSTEM) [communicator] → produces: confirm_id, capabilities: produces_confirm,can_succeed,outcome_success\n 4. UPDATE_RECORDS_V11 (SYSTEM) [processor] → produces: record_updated, capabilities: produces_record,can_succeed,outcome_success\n\nState:\n goal_progress=0.14\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=0.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "SEND_CONFIRMATION_V11"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: APPOINTMENT_BOOKING (AGENT)\nOutcome: customer_declined\n\nState:\n goal_progress=0.36\n retry_count=1\n total_retries=1\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=2\n nodes_visited=2\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=32.00\n cost_accrued=20.50\n uncertainty_level=0.60\n\nEligible nodes:\n 1. APPOINTMENT_REMINDERS (SYSTEM) [coordinator] → produces: appointment_id intent:scheduling,appointment,coordination\n 2. END_CUSTOMER_DECLINED (SYSTEM) [terminator] → produces: none intent:rejection,termination,completion\n 3. APPOINTMENT_BOOKING (AGENT) [retry_handler] → produces: appointment_id, appointment_datetime, concerns intent:scheduling,appointment,coordination\n 4. END_ESCALATED_ISSUE (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n\nForkable sets: [{END_CUSTOMER_DECLINED, END_ESCALATED_ISSUE}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "FORK"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: BRANCH_2 (AGENT)\nOutcome: failure\n\nState:\n goal_progress=0.50\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=1\n nodes_visited=2\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.07\n elapsed_time_minutes=17.00\n cost_accrued=10.50\n uncertainty_level=0.35\n time_to_deadline_minutes=223\n\nEligible nodes:\n 1. BRANCH_2 (AGENT) [retry_handler] → produces: ctx_b2\n 2. JOIN (SYSTEM) [processor] → produces: ctx_joined intent:automated\n 3. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_FAIL, JOIN}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "FORK"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_1 (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=0.17\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=0\n nodes_visited=1\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=2.00\n cost_accrued=0.50\n uncertainty_level=0.15\n\nEligible nodes:\n 1. STEP_2 (AGENT) [processor] → produces: ctx_step2\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n\nForkable sets: [{END_ESCALATED, STEP_2}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: BRANCH_4_STEP_2 (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=0.27\n retry_count=0\n total_retries=0\n parallel_active=4\n parallel_depth=1\n parallel_completed=1\n steps_taken=3\n nodes_visited=6\n escalation_count=0\n resource_pressure=0.60\n sla_pressure=0.01\n elapsed_time_minutes=8.00\n cost_accrued=2.00\n uncertainty_level=0.00\n time_to_deadline_minutes=1432\n\nEligible nodes:\n 1. JOIN_POINT (SYSTEM) [processor] → produces: ctx_joined intent:automated\n 2. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_FAIL, JOIN_POINT}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_3 (HUMAN)\nOutcome: success\n\nState:\n goal_progress=0.60\n retry_count=0\n total_retries=0\n parallel_active=4\n parallel_depth=2\n parallel_completed=1\n steps_taken=4\n nodes_visited=4\n escalation_count=0\n resource_pressure=0.60\n sla_pressure=0.04\n elapsed_time_minutes=51.00\n cost_accrued=36.50\n uncertainty_level=0.00\n time_to_deadline_minutes=1389\n\nEligible nodes:\n 1. STEP_4 (SYSTEM) [processor] → produces: ctx_step4 intent:automated\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_3 (HUMAN)\nOutcome: failure\n\nState:\n goal_progress=1.00\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=2\n nodes_visited=3\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=47.00\n cost_accrued=35.50\n uncertainty_level=0.70\n\nEligible nodes:\n 1. STEP_3 (HUMAN) [retry_handler] → produces: ctx_step_3 intent:human_task\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n 3. END_SUCCESS (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_ESCALATED, END_SUCCESS}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "FORK"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type FORK, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: FORK\n\nCandidates:\n 1. STEP_2 (AGENT) [processor] → produces: ctx_step_2, capabilities: produces_ctx,can_succeed,outcome_success\n 2. STEP_1 (SYSTEM) [retry_handler] → produces: ctx_step_1, capabilities: produces_ctx,can_succeed,outcome_success\n 3. END_ESCALATED (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n 4. END_ESCALATED_STEP_1 (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=0.33\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.20\n cost_accrued=0.50\n\nSelect nodes to fork:"
},
{
"role": "assistant",
"content": "END_ESCALATED, END_ESCALATED_STEP_1, STEP_2"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type JOIN, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: JOIN\n\nCandidates:\n 1. STEP_2 (AGENT) [retry_handler] → produces: ctx_step_2, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.67\n parallel_depth=1\n retry_count=0\n sla_pressure=0.13\n resource_pressure=0.30\n uncertainty_level=0.30\n cost_accrued=20.50\n\nSelect join node:"
},
{
"role": "assistant",
"content": "STEP_2"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step_3, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.67\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.10\n cost_accrued=10.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "STEP_3"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type RETRY, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: RETRY\n\nCandidates:\n 1. STAGE_2 (AGENT) [retry_handler] → produces: ctx_stage2, capabilities: produces_ctx,can_succeed,outcome_success\n 2. STAGE_3 (HUMAN) [processor] → produces: ctx_stage3, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.20\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.30\n cost_accrued=10.50\n\nSelect node for RETRY:"
},
{
"role": "assistant",
"content": "STAGE_2"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. STAGE_2 (AGENT) [processor] → produces: ctx_stage2, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.12\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=0.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "STAGE_2"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: BRANCH_2 (AGENT)\nOutcome: retry\n\nState:\n goal_progress=0.40\n retry_count=0\n total_retries=1\n parallel_active=2\n parallel_depth=0\n parallel_completed=0\n steps_taken=8\n nodes_visited=4\n escalation_count=0\n resource_pressure=0.30\n sla_pressure=0.51\n elapsed_time_minutes=122.00\n cost_accrued=80.50\n uncertainty_level=0.10\n time_to_deadline_minutes=118\n\nEligible nodes:\n 1. BRANCH_2 (AGENT) [retry_handler] → produces: ctx_b2\n 2. JOIN (SYSTEM) [processor] → produces: ctx_joined intent:automated\n 3. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_FAIL, JOIN}]\nJoin-ready: ['BRANCH_2']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "RETRY"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. STEP_2 (AGENT) [processor] → produces: ctx_step_2, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.33\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=0.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "STEP_2"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: INTAKE (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=0.17\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=0\n nodes_visited=0\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=0.00\n cost_accrued=0.00\n uncertainty_level=0.00\n\nEligible nodes:\n 1. NOTIFY_CUSTOMER_V8 (SYSTEM) → produces: none\n 2. SEND_CONFIRMATION_V8 (SYSTEM) → produces: none\n 3. LOG_AUDIT_TRAIL_V8 (SYSTEM) → produces: none\n 4. END_FAIL (SYSTEM) → produces: none\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "META"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: TASK_1 (SYSTEM)\nOutcome: failure\n\nState:\n goal_progress=0.50\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=1\n nodes_visited=2\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=4.00\n cost_accrued=1.00\n uncertainty_level=0.25\n\nEligible nodes:\n 1. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: FINALIZE (AGENT)\nOutcome: retry\n\nState:\n goal_progress=0.67\n retry_count=0\n total_retries=0\n parallel_active=2\n parallel_depth=0\n parallel_completed=0\n steps_taken=10\n nodes_visited=6\n escalation_count=0\n resource_pressure=0.30\n sla_pressure=0.00\n elapsed_time_minutes=126.00\n cost_accrued=81.50\n uncertainty_level=0.30\n\nEligible nodes:\n 1. END_SUCCESS (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n 2. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n 3. FINALIZE (AGENT) [retry_handler] → produces: ctx_final\n\nForkable sets: [{END_FAIL, END_SUCCESS}]\nJoin-ready: ['FINALIZE']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "JOIN"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_2 (SYSTEM)\nOutcome: authorization_failure\n\nState:\n goal_progress=0.50\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=0\n nodes_visited=0\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=0.00\n cost_accrued=0.00\n uncertainty_level=0.00\n\nEligible nodes:\n 1. END_ESCALATED (SYSTEM) → produces: none\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "META"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: PATH_2 (AGENT)\nOutcome: success\n\nState:\n goal_progress=0.50\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=1\n nodes_visited=2\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.07\n elapsed_time_minutes=17.00\n cost_accrued=10.50\n uncertainty_level=0.20\n time_to_deadline_minutes=223\n\nEligible nodes:\n 1. CONVERGE (SYSTEM) [processor] → produces: ctx_done intent:automated\n 2. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{CONVERGE, END_FAIL}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: BRANCH_3 (HUMAN)\nOutcome: retry\n\nState:\n goal_progress=0.40\n retry_count=1\n total_retries=1\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=2\n nodes_visited=2\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=62.00\n cost_accrued=50.50\n uncertainty_level=0.30\n\nEligible nodes:\n 1. BRANCH_3 (HUMAN) [retry_handler] → produces: ctx_b3 intent:human_task\n 2. JOIN (SYSTEM) [processor] → produces: ctx_joined intent:automated\n 3. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_FAIL, JOIN}]\nJoin-ready: ['BRANCH_3']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "RETRY"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: BRANCH_2 (AGENT)\nOutcome: failure\n\nState:\n goal_progress=0.50\n retry_count=1\n total_retries=1\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=2\n nodes_visited=2\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=32.00\n cost_accrued=20.50\n uncertainty_level=0.25\n\nEligible nodes:\n 1. BRANCH_2 (AGENT) [retry_handler] → produces: ctx_b2\n 2. JOIN (SYSTEM) [processor] → produces: ctx_joined intent:automated\n 3. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_FAIL, JOIN}]\nJoin-ready: ['BRANCH_2']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "FORK"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step_3, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.33\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.40\n cost_accrued=10.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "STEP_3"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: CONVERGE (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=0.50\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=2\n nodes_visited=3\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=6.00\n cost_accrued=1.50\n uncertainty_level=0.05\n\nEligible nodes:\n 1. END_SUCCESS (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: INITIAL_TRIAGE (AGENT)\nOutcome: insufficient_info\n\nState:\n goal_progress=0.23\n retry_count=0\n total_retries=0\n parallel_active=3\n parallel_depth=0\n parallel_completed=0\n steps_taken=78\n nodes_visited=5\n escalation_count=0\n resource_pressure=0.45\n sla_pressure=1.00\n elapsed_time_minutes=1172.00\n cost_accrued=780.50\n uncertainty_level=0.00\n time_to_deadline_minutes=0\n\nEligible nodes:\n 1. INITIAL_TRIAGE (AGENT) [processor] → produces: severity_level, claim_category, estimated_amount\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n\nForkable sets: []\nJoin-ready: ['INITIAL_TRIAGE']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "JOIN"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: INITIAL_TRIAGE (AGENT)\nOutcome: insufficient_info\n\nState:\n goal_progress=0.23\n retry_count=0\n total_retries=0\n parallel_active=3\n parallel_depth=0\n parallel_completed=0\n steps_taken=51\n nodes_visited=5\n escalation_count=0\n resource_pressure=0.45\n sla_pressure=1.00\n elapsed_time_minutes=754.00\n cost_accrued=501.00\n uncertainty_level=0.00\n time_to_deadline_minutes=0\n\nEligible nodes:\n 1. INITIAL_TRIAGE (AGENT) [processor] → produces: severity_level, claim_category, estimated_amount\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n\nForkable sets: []\nJoin-ready: ['INITIAL_TRIAGE']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "JOIN"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: FINALIZE (AGENT)\nOutcome: success\n\nState:\n goal_progress=0.57\n retry_count=1\n total_retries=1\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=4\n nodes_visited=4\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=44.00\n cost_accrued=26.00\n uncertainty_level=0.05\n\nEligible nodes:\n 1. END_SUCCESS (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: INITIAL_TRIAGE (AGENT)\nOutcome: insufficient_info\n\nState:\n goal_progress=0.23\n retry_count=0\n total_retries=0\n parallel_active=3\n parallel_depth=0\n parallel_completed=0\n steps_taken=6\n nodes_visited=5\n escalation_count=0\n resource_pressure=0.45\n sla_pressure=0.00\n elapsed_time_minutes=92.00\n cost_accrued=60.50\n uncertainty_level=0.00\n\nEligible nodes:\n 1. INITIAL_TRIAGE (AGENT) [processor] → produces: severity_level, claim_category, estimated_amount\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n\nForkable sets: []\nJoin-ready: ['INITIAL_TRIAGE']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "JOIN"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step_3, capabilities: produces_ctx,can_succeed,outcome_success\n 2. STEP_2 (AGENT) [retry_handler] → produces: ctx_step_2, capabilities: produces_ctx,can_succeed,outcome_success\n 3. END_ESCALATED (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=0.40\n parallel_depth=0\n retry_count=3\n sla_pressure=0.00\n resource_pressure=0.30\n uncertainty_level=0.60\n cost_accrued=70.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "END_ESCALATED"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type FORK, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: FORK\n\nCandidates:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step_3, capabilities: produces_ctx,can_succeed,outcome_success\n 2. STEP_2 (AGENT) [retry_handler] → produces: ctx_step_2, capabilities: produces_ctx,can_succeed,outcome_success\n 3. END_ESCALATED (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=0.67\n parallel_depth=0\n retry_count=0\n sla_pressure=0.07\n resource_pressure=0.00\n uncertainty_level=0.35\n cost_accrued=10.50\n\nSelect nodes to fork:"
},
{
"role": "assistant",
"content": "END_ESCALATED, STEP_3"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_5 (AGENT)\nOutcome: retry\n\nState:\n goal_progress=1.00\n retry_count=1\n total_retries=1\n parallel_active=2\n parallel_depth=1\n parallel_completed=0\n steps_taken=6\n nodes_visited=7\n escalation_count=0\n resource_pressure=0.30\n sla_pressure=0.00\n elapsed_time_minutes=94.00\n cost_accrued=66.00\n uncertainty_level=0.05\n\nEligible nodes:\n 1. STEP_5 (AGENT) [retry_handler] → produces: ctx_step_5\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n 3. END_SUCCESS (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_ESCALATED, END_SUCCESS}]\nJoin-ready: ['STEP_5']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "RETRY"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_2 (AGENT)\nOutcome: failure\n\nState:\n goal_progress=0.29\n retry_count=0\n total_retries=0\n parallel_active=2\n parallel_depth=1\n parallel_completed=0\n steps_taken=2\n nodes_visited=4\n escalation_count=0\n resource_pressure=0.30\n sla_pressure=0.00\n elapsed_time_minutes=32.00\n cost_accrued=20.50\n uncertainty_level=0.55\n\nEligible nodes:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step_3 intent:human_task\n 2. END_EXIT_2 (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_EXIT_2, STEP_3}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "FORK"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type RETRY, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: RETRY\n\nCandidates:\n 1. BRANCH_1 (SYSTEM) [retry_handler] → produces: ctx_b1, capabilities: produces_ctx,can_succeed,outcome_success\n 2. JOIN (SYSTEM) [processor] → produces: ctx_joined, capabilities: produces_ctx,can_succeed,outcome_success\n 3. END_FAIL (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=0.50\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.10\n cost_accrued=1.00\n\nSelect node for RETRY:"
},
{
"role": "assistant",
"content": "BRANCH_1"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. END_SUCCESS (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=1.00\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.00\n cost_accrued=46.00\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "END_SUCCESS"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. SEND_THANK_YOU_SMS (SYSTEM) [communicator] → produces: customer_id, capabilities: produces_customer,can_succeed,outcome_sms_sent\n 2. END_PAYMENT_FAILURE (SYSTEM) [terminator] → produces: none\n\nState:\n goal_progress=0.82\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.10\n cost_accrued=166.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "SEND_THANK_YOU_SMS"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_6 (HUMAN)\nOutcome: retry\n\nState:\n goal_progress=1.00\n retry_count=1\n total_retries=1\n parallel_active=2\n parallel_depth=1\n parallel_completed=0\n steps_taken=7\n nodes_visited=8\n escalation_count=0\n resource_pressure=0.30\n sla_pressure=0.00\n elapsed_time_minutes=154.00\n cost_accrued=121.00\n uncertainty_level=0.20\n\nEligible nodes:\n 1. STEP_6 (HUMAN) [retry_handler] → produces: ctx_step6 intent:human_task\n 2. END_SUCCESS (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n 3. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_FAIL, END_SUCCESS}]\nJoin-ready: ['STEP_6']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "JOIN"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type JOIN, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: JOIN\n\nCandidates:\n 1. BRANCH_2 (AGENT) [retry_handler] → produces: ctx_b2, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.40\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.30\n uncertainty_level=0.40\n cost_accrued=80.50\n\nSelect join node:"
},
{
"role": "assistant",
"content": "BRANCH_2"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type FORK, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: FORK\n\nCandidates:\n 1. STEP_5 (AGENT) [retry_handler] → produces: ctx_step_5, capabilities: produces_ctx,can_succeed,outcome_success\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n 3. END_SUCCESS (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=1.00\n parallel_depth=0\n retry_count=1\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=56.00\n\nSelect nodes to fork:"
},
{
"role": "assistant",
"content": "END_ESCALATED, END_SUCCESS"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. VEHICLE_RECEPTION (HUMAN) [processor] → produces: vehicle_id, customer_id, sla: 10min, capabilities: produces_vehicle,produces_customer,can_succeed\n\nState:\n goal_progress=0.36\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.00\n cost_accrued=11.00\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "VEHICLE_RECEPTION"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. PATH_A (SYSTEM) [processor] → produces: ctx_patha, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.33\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=0.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "PATH_A"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type JOIN, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: JOIN\n\nCandidates:\n 1. FINALIZE (AGENT) [retry_handler] → produces: ctx_final, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.57\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.30\n uncertainty_level=0.00\n cost_accrued=186.00\n\nSelect join node:"
},
{
"role": "assistant",
"content": "FINALIZE"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: BRANCH_2 (AGENT)\nOutcome: retry\n\nState:\n goal_progress=0.50\n retry_count=0\n total_retries=1\n parallel_active=2\n parallel_depth=0\n parallel_completed=0\n steps_taken=11\n nodes_visited=4\n escalation_count=0\n resource_pressure=0.30\n sla_pressure=0.00\n elapsed_time_minutes=167.00\n cost_accrued=110.50\n uncertainty_level=0.30\n\nEligible nodes:\n 1. BRANCH_2 (AGENT) [retry_handler] → produces: ctx_b2\n 2. JOIN (SYSTEM) [processor] → produces: ctx_joined intent:automated\n 3. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_FAIL, JOIN}]\nJoin-ready: ['BRANCH_2']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "JOIN"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type RETRY, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: RETRY\n\nCandidates:\n 1. INSPECT_CONDITION_V3 (HUMAN) [retry_handler] → produces: condition_report, capabilities: produces_condition,can_succeed,outcome_success\n 2. APPROVE_WORK_V3 (AGENT) [gatekeeper] → produces: approval_status, capabilities: produces_approval,can_succeed,outcome_success\n 3. END_FAIL (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=0.40\n parallel_depth=0\n retry_count=1\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.10\n cost_accrued=50.50\n\nSelect node for RETRY:"
},
{
"role": "assistant",
"content": "INSPECT_CONDITION_V3"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type RETRY, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: RETRY\n\nCandidates:\n 1. BRANCH_1_STEP_2 (SYSTEM) [processor] → produces: ctx_b1_s2, capabilities: produces_ctx,can_succeed,outcome_success\n 2. BRANCH_1_STEP_1 (SYSTEM) [retry_handler] → produces: ctx_b1_s1, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.18\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.60\n cost_accrued=1.00\n\nSelect node for RETRY:"
},
{
"role": "assistant",
"content": "BRANCH_1_STEP_1"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type JOIN, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: JOIN\n\nCandidates:\n 1. INITIAL_TRIAGE (AGENT) [processor] → produces: severity_level, claim_category, estimated_amount, capabilities: produces_severity,produces_claim,produces_estimated\n\nState:\n goal_progress=0.23\n parallel_depth=0\n retry_count=0\n sla_pressure=0.07\n resource_pressure=0.45\n uncertainty_level=0.00\n cost_accrued=70.50\n\nSelect join node:"
},
{
"role": "assistant",
"content": "INITIAL_TRIAGE"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. STEP_4 (SYSTEM) [retry_handler] → produces: ctx_step4, capabilities: produces_ctx,can_succeed,outcome_success\n 2. STEP_5 (AGENT) [processor] → produces: ctx_step5, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.67\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.10\n cost_accrued=36.00\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "STEP_5"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. SENSOR_READING_V10 (DEVICE) [processor] → produces: sensor_data, capabilities: produces_sensor,can_succeed,outcome_success\n 2. AUTOMATED_TEST_V10 (SYSTEM) [processor] → produces: test_result, capabilities: produces_test,can_succeed,outcome_success\n 3. QUALITY_CHECK_V10 (HUMAN) [validator] → produces: quality_score, capabilities: produces_quality,can_succeed,outcome_success\n 4. PHYSICAL_INSPECTION_V10 (HUMAN) [validator] → produces: inspection_result, capabilities: produces_inspection,can_succeed,outcome_success\n 5. END_FAIL (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=0.14\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=0.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "AUTOMATED_TEST_V10"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type JOIN, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: JOIN\n\nCandidates:\n 1. FINALIZE (AGENT) [retry_handler] → produces: ctx_final, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.67\n parallel_depth=0\n retry_count=2\n sla_pressure=0.59\n resource_pressure=0.30\n uncertainty_level=0.10\n cost_accrued=91.50\n\nSelect join node:"
},
{
"role": "assistant",
"content": "FINALIZE"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type JOIN, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: JOIN\n\nCandidates:\n 1. BRANCH_2 (AGENT) [retry_handler] → produces: ctx_b2, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.40\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.30\n uncertainty_level=0.55\n cost_accrued=50.50\n\nSelect join node:"
},
{
"role": "assistant",
"content": "BRANCH_2"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step3, capabilities: produces_ctx,can_succeed,outcome_success\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=0.33\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.20\n cost_accrued=10.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "STEP_3"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type RETRY, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: RETRY\n\nCandidates:\n 1. STEP_5 (AGENT) [retry_handler] → produces: ctx_step_5, capabilities: produces_ctx,can_succeed,outcome_success\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n 3. END_SUCCESS (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=1.00\n parallel_depth=1\n retry_count=0\n sla_pressure=0.33\n resource_pressure=0.30\n uncertainty_level=0.15\n cost_accrued=56.00\n\nSelect node for RETRY:"
},
{
"role": "assistant",
"content": "STEP_5"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type FORK, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: FORK\n\nCandidates:\n 1. BRANCH_1_STEP_1 (SYSTEM) [processor] → produces: ctx_b1_s1, capabilities: produces_ctx,can_succeed,outcome_success\n 2. BRANCH_2_STEP_1 (AGENT) [processor] → produces: ctx_b2_s1, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.14\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=0.50\n\nSelect nodes to fork:"
},
{
"role": "assistant",
"content": "BRANCH_1_STEP_1, BRANCH_2_STEP_1"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: INITIAL_TRIAGE (AGENT)\nOutcome: insufficient_info\n\nState:\n goal_progress=0.23\n retry_count=0\n total_retries=0\n parallel_active=3\n parallel_depth=0\n parallel_completed=0\n steps_taken=52\n nodes_visited=5\n escalation_count=0\n resource_pressure=0.45\n sla_pressure=1.00\n elapsed_time_minutes=782.00\n cost_accrued=520.50\n uncertainty_level=0.00\n time_to_deadline_minutes=0\n\nEligible nodes:\n 1. INITIAL_TRIAGE (AGENT) [processor] → produces: severity_level, claim_category, estimated_amount\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n\nForkable sets: []\nJoin-ready: ['INITIAL_TRIAGE']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "JOIN"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. AUTOMATED_ASSESSMENT_V15 (SYSTEM) [processor] → produces: auto_score, capabilities: produces_auto,can_succeed,outcome_success\n 2. MANUAL_REVIEW_V15 (HUMAN) [validator] → produces: manual_score, capabilities: produces_manual,can_succeed,outcome_success\n\nState:\n goal_progress=0.20\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.25\n cost_accrued=0.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "MANUAL_REVIEW_V15"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STAGE_6 (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=0.86\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=0\n nodes_visited=0\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=0.00\n cost_accrued=0.00\n uncertainty_level=0.00\n\nEligible nodes:\n 1. STAGE_7 (SYSTEM) → produces: none\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "META"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step3, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.40\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.00\n cost_accrued=30.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "STEP_3"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. AUTOMATED_TEST_V17 (SYSTEM) [processor] → produces: test_result, capabilities: produces_test,can_succeed,outcome_success\n 2. REMOTE_DIAGNOSTICS_V17 (DEVICE) [processor] → produces: diag_result, capabilities: produces_diag,can_succeed,outcome_success\n 3. SENSOR_READING_V17 (DEVICE) [processor] → produces: sensor_data, capabilities: produces_sensor,can_succeed,outcome_success\n 4. QUALITY_CHECK_V17 (HUMAN) [validator] → produces: quality_score, capabilities: produces_quality,can_succeed,outcome_success\n 5. END_FAIL (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=0.14\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=0.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "REMOTE_DIAGNOSTICS_V17"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type JOIN, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: JOIN\n\nCandidates:\n 1. INITIAL_TRIAGE (AGENT) [processor] → produces: severity_level, claim_category, estimated_amount, capabilities: produces_severity,produces_claim,produces_estimated\n\nState:\n goal_progress=0.23\n parallel_depth=0\n retry_count=0\n sla_pressure=0.51\n resource_pressure=0.45\n uncertainty_level=0.00\n cost_accrued=80.50\n\nSelect join node:"
},
{
"role": "assistant",
"content": "INITIAL_TRIAGE"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. END_SUCCESS (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n 2. END_FAIL (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=1.00\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.05\n cost_accrued=127.00\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "END_SUCCESS"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: COMPLETE (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=0.80\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=3\n nodes_visited=4\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=49.00\n cost_accrued=36.00\n uncertainty_level=0.00\n\nEligible nodes:\n 1. END_SUCCESS (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. STEP_2 (AGENT) [processor] → produces: ctx_step2, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.20\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=0.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "STEP_2"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_1 (SYSTEM)\nOutcome: data_corruption_detected\n\nState:\n goal_progress=0.20\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=0\n nodes_visited=1\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=2.00\n cost_accrued=0.50\n uncertainty_level=0.05\n time_to_deadline_minutes=1438\n\nEligible nodes:\n 1. STEP_2 (AGENT) [processor] → produces: ctx_step2\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n\nForkable sets: [{END_ESCALATED, STEP_2}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "FORK"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: INITIAL_TRIAGE (AGENT)\nOutcome: insufficient_info\n\nState:\n goal_progress=0.23\n retry_count=0\n total_retries=0\n parallel_active=3\n parallel_depth=0\n parallel_completed=0\n steps_taken=47\n nodes_visited=5\n escalation_count=0\n resource_pressure=0.45\n sla_pressure=0.49\n elapsed_time_minutes=707.00\n cost_accrued=470.50\n uncertainty_level=0.00\n time_to_deadline_minutes=733\n\nEligible nodes:\n 1. INITIAL_TRIAGE (AGENT) [processor] → produces: severity_level, claim_category, estimated_amount\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n\nForkable sets: []\nJoin-ready: ['INITIAL_TRIAGE']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "JOIN"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: BRANCH_3 (HUMAN)\nOutcome: retry\n\nState:\n goal_progress=0.40\n retry_count=0\n total_retries=0\n parallel_active=2\n parallel_depth=1\n parallel_completed=0\n steps_taken=2\n nodes_visited=4\n escalation_count=0\n resource_pressure=0.30\n sla_pressure=0.00\n elapsed_time_minutes=62.00\n cost_accrued=50.50\n uncertainty_level=0.20\n\nEligible nodes:\n 1. BRANCH_3 (HUMAN) [retry_handler] → produces: ctx_b3 intent:human_task\n 2. JOIN (SYSTEM) [processor] → produces: ctx_joined intent:automated\n 3. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_FAIL, JOIN}]\nJoin-ready: ['BRANCH_3']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "JOIN"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. END_FAIL (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=0.57\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=36.00\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "END_FAIL"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_1 (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=0.33\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=0\n nodes_visited=1\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=2.00\n cost_accrued=0.50\n uncertainty_level=0.65\n\nEligible nodes:\n 1. STEP_2 (AGENT) [processor] → produces: ctx_step2\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n\nForkable sets: [{END_ESCALATED, STEP_2}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: PATH_B (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=0.67\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=0\n nodes_visited=0\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=0.00\n cost_accrued=0.00\n uncertainty_level=0.00\n\nEligible nodes:\n 1. END_B (SYSTEM) → produces: none\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "META"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. TASK_1 (SYSTEM) [processor] → produces: ctx_task1, capabilities: produces_ctx,can_succeed,outcome_success\n 2. TASK_2 (AGENT) [processor] → produces: ctx_task2, capabilities: produces_ctx,can_succeed,outcome_success\n 3. TASK_3 (HUMAN) [processor] → produces: ctx_task3, capabilities: produces_ctx,can_succeed,outcome_success\n 4. TASK_4 (SYSTEM) [processor] → produces: ctx_task4, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.17\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.05\n cost_accrued=0.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "TASK_3"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. BRANCH_1_STEP_1 (SYSTEM) [processor] → produces: ctx_b1_s1, capabilities: produces_ctx,can_succeed,outcome_success\n 2. BRANCH_2_STEP_1 (AGENT) [processor] → produces: ctx_b2_s1, capabilities: produces_ctx,can_succeed,outcome_success\n 3. BRANCH_3_STEP_1 (HUMAN) [processor] → produces: ctx_b3_s1, capabilities: produces_ctx,can_succeed,outcome_success\n 4. BRANCH_4_STEP_1 (SYSTEM) [processor] → produces: ctx_b4_s1, capabilities: produces_ctx,can_succeed,outcome_success\n\nState:\n goal_progress=0.09\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=0.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "BRANCH_1_STEP_1"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type NEXT, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: NEXT\n\nCandidates:\n 1. DIAGNOSIS_CONCLUSION_V10 (HUMAN) [processor] → produces: diagnosis_result, capabilities: produces_diagnosis,can_succeed,outcome_success\n\nState:\n goal_progress=0.33\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.10\n cost_accrued=10.50\n\nSelect node for NEXT:"
},
{
"role": "assistant",
"content": "DIAGNOSIS_CONCLUSION_V10"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the decision type FORK, select the node(s) to execute from the candidates."
},
{
"role": "user",
"content": "Decision: FORK\n\nCandidates:\n 1. STEP_5 (AGENT) [processor] → produces: ctx_step_5, capabilities: produces_ctx,can_succeed,outcome_success\n 2. STEP_4 (SYSTEM) [retry_handler] → produces: ctx_step_4, capabilities: produces_ctx,can_succeed,outcome_success\n 3. END_ESCALATED (SYSTEM) [terminator] → produces: none, capabilities: can_succeed,outcome_success,can_fail\n\nState:\n goal_progress=0.80\n parallel_depth=0\n retry_count=0\n sla_pressure=0.00\n resource_pressure=0.00\n uncertainty_level=0.15\n cost_accrued=36.00\n\nSelect nodes to fork:"
},
{
"role": "assistant",
"content": "END_ESCALATED, STEP_5"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_2 (AGENT)\nOutcome: success\n\nState:\n goal_progress=0.40\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=1\n nodes_visited=2\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=17.00\n cost_accrued=10.50\n uncertainty_level=0.10\n\nEligible nodes:\n 1. STEP_3 (HUMAN) [processor] → produces: ctx_step3 intent:human_task\n\nForkable sets: []\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: STEP_1 (SYSTEM)\nOutcome: external_dependency_missing\n\nState:\n goal_progress=0.17\n retry_count=0\n total_retries=0\n parallel_active=2\n parallel_depth=1\n parallel_completed=0\n steps_taken=1\n nodes_visited=3\n escalation_count=0\n resource_pressure=0.30\n sla_pressure=0.00\n elapsed_time_minutes=4.00\n cost_accrued=1.00\n uncertainty_level=0.00\n time_to_deadline_minutes=1436\n\nEligible nodes:\n 1. STEP_2 (AGENT) [processor] → produces: ctx_step2\n 2. END_ESCALATED (SYSTEM) [terminator] → produces: none intent:escalation,exception_handling,supervisor\n\nForkable sets: [{END_ESCALATED, STEP_2}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "FORK"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: BRANCH_3 (HUMAN)\nOutcome: retry\n\nState:\n goal_progress=0.40\n retry_count=0\n total_retries=1\n parallel_active=2\n parallel_depth=0\n parallel_completed=0\n steps_taken=7\n nodes_visited=4\n escalation_count=0\n resource_pressure=0.30\n sla_pressure=0.00\n elapsed_time_minutes=212.00\n cost_accrued=175.50\n uncertainty_level=0.50\n\nEligible nodes:\n 1. BRANCH_3 (HUMAN) [retry_handler] → produces: ctx_b3 intent:human_task\n 2. JOIN (SYSTEM) [processor] → produces: ctx_joined intent:automated\n 3. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_FAIL, JOIN}]\nJoin-ready: ['BRANCH_3']\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "JOIN"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: START (SYSTEM)\nOutcome: success\n\nState:\n goal_progress=0.20\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=0\n nodes_visited=1\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=2.00\n cost_accrued=0.50\n uncertainty_level=0.35\n\nEligible nodes:\n 1. ESTIMATE_COST_V13 (AGENT) [processor] → produces: estimate_amount\n 2. INSPECT_CONDITION_V13 (HUMAN) [validator] → produces: condition_report intent:validation,quality_check,human_task\n\nForkable sets: [{ESTIMATE_COST_V13, INSPECT_CONDITION_V13}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "NEXT"
}
] |
[
{
"role": "system",
"content": "You are a workflow planner. Given the current workflow state, eligible nodes, and topology information, classify the decision type."
},
{
"role": "user",
"content": "Current node: BRANCH_2 (AGENT)\nOutcome: failure\n\nState:\n goal_progress=0.50\n retry_count=0\n total_retries=0\n parallel_active=0\n parallel_depth=0\n parallel_completed=0\n steps_taken=1\n nodes_visited=2\n escalation_count=0\n resource_pressure=0.00\n sla_pressure=0.00\n elapsed_time_minutes=17.00\n cost_accrued=10.50\n uncertainty_level=0.75\n\nEligible nodes:\n 1. BRANCH_2 (AGENT) [retry_handler] → produces: ctx_b2\n 2. JOIN (SYSTEM) [processor] → produces: ctx_joined intent:automated\n 3. END_FAIL (SYSTEM) [terminator] → produces: none intent:termination,completion,automated\n\nForkable sets: [{END_FAIL, JOIN}]\nJoin-ready: []\n\nWhat decision type?"
},
{
"role": "assistant",
"content": "FORK"
}
] |
End of preview. Expand
in Data Studio
SLM Workflow Planner — Policy-Corrected Instruction Tuning Dataset (v2)
Overview
High-quality instruction-tuning dataset for training a Small Language Model (SLM) to serve as a workflow execution planner. The model learns to make policy-aware decisions about workflow transitions: when to proceed (NEXT), retry (RETRY), parallelize (FORK), synchronize (JOIN), or escalate (META).
Key Features
- 648K instruction pairs across 2 stages (decision type + node selection)
- Policy-corrected labels: FORK/JOIN/META decisions are conditioned on state signals (resource pressure, parallel activity, goal progress, uncertainty), not just topology flags
- Counterfactual negatives: Scenarios where topology suggests FORK/JOIN but state signals dictate NEXT — teaching the model true policy boundaries
- 89 diverse workflow graphs across 8 structural families + 2 semantic workflows
- Balanced decision distribution: NEXT ~36%, JOIN ~27%, META ~13%, FORK ~12%, RETRY ~12%
Dataset Structure
Each sample is a chat-completion format message with 3 roles:
{
"messages": [
{"role": "system", "content": "You are a workflow planner..."},
{"role": "user", "content": "Current node: ... State: ... Eligible: ..."},
{"role": "assistant", "content": "FORK"}
]
}
Stage 1: Decision Type Classification
Given current node, outcome, state signals, eligible nodes, forkable sets, and
join-ready nodes → predict one of: NEXT, RETRY, FORK, JOIN, META
Stage 2: Node Selection
Given the decision type and candidates → select which node(s) to execute
State Signals
Each sample includes realistic state signals:
goal_progress(0→1): How close to workflow completionretry_count/total_retries: Retry budget trackingparallel_active/parallel_depth: Current parallelism stateresource_pressure(0→1): System load indicatorsla_pressure(0→1): Deadline urgencyuncertainty_level(0→1): Confidence in current pathcost_accrued: Cumulative cost so farescalation_count: Number of escalations triggered
Policy Boundaries (What Makes This Dataset Special)
Unlike naive topology-based labeling, this dataset includes:
| Scenario | Topology says | State says | Label |
|---|---|---|---|
| Forkable set present + high resource pressure | FORK | Don't parallelize | NEXT |
| Forkable set present + already parallel | FORK | Already forked | NEXT |
| Join-ready + low goal progress | JOIN | Too early | NEXT |
| Anomaly outcome + high uncertainty | NEXT | Escalate | META |
These counterfactual samples are critical for learning policy vs topology.
Workflow Families
Generated from 89 workflows spanning:
- Linear with optional exits
- Retry loops
- Fork-Join patterns
- OR-alternative paths
- Escalation ladders
- Partial (k-of-n) joins
- Conditional irreversible branches
- Long horizon workflows
- Semantic parallel (Insurance Claim)
- Semantic join patterns
Intended Use
Fine-tuning SLMs (3B–7B) with LoRA for workflow planning tasks.
Tested with: Qwen/Qwen2.5-7B-Instruct
Splits
| Split | Samples |
|---|---|
| Train | 583,504 |
| Valid | 32,417 |
| Test | 32,417 |
| Total | 648,338 |
Citation
Part of the Agentic Factory project — building autonomous workflow orchestration with SLM-powered planning.
- Downloads last month
- 12
Size of downloaded dataset files:
516 MB
Size of the auto-converted Parquet files:
449 MB
Number of rows:
648,338
Models trained or fine-tuned on ssaraf1/slm-workflow-planner-policy-v2
