By name: rolling_update_fleet description: >- Parallel execution workflow for rolling update fleet using the Unified Parallel Engine domain: system agent: systems_engineer team_config: name: systems_operations_team task_pattern: system administration and management execution_mode: parallel specialist_ids: - scanner-agent - analyzer-agent - remediator-agent - reporter-agent tool_assignments: scanner-agent: [tun_tm_system, tun_tm_remote] analyzer-agent: [graph_analyze, tun_tm_security] remediator-agent: [tun_tm_remote, tun_tm_inventory] reporter-agent: [graph_write, document_tools] tags: [system, rolling-update-fleet]
concept: CONCEPT:SYS-001
Rolling Update Fleet Workflow
CONCEPT:SYS-001
Parallel execution workflow for rolling update fleet using the Unified Parallel Engine
Steps
Step 1: Wave Per Batch Drain
Agent: scanner-agent
Tools: tun_tm_system, tun_tm_remote
Execute wave per batch drain operations for the Rolling Update Fleet workflow.
Expected: wave_per_batch_drain_artifacts
Step 2: Update [depends_on: wave_per_batch_drain]
Agent: analyzer-agent
Tools: graph_analyze, tun_tm_security
Execute update operations for the Rolling Update Fleet workflow.
Expected: update_artifacts
Step 3: Health Check [depends_on: update]
Agent: remediator-agent
Tools: tun_tm_remote, tun_tm_inventory
Execute health check operations for the Rolling Update Fleet workflow.
Expected: health_check_artifacts
Step 4: Next Batch [depends_on: health_check]
Agent: reporter-agent
Tools: graph_write, document_tools
Execute next batch operations for the Rolling Update Fleet workflow.
Expected: next_batch_artifacts
Step 5: KG Persistence [depends_on: next_batch]
Agent: reporter-agent
Tools: graph_write
Persist workflow results as nodes and edges in the Knowledge Graph. Create appropriate typed nodes with metadata and link to existing domain entities.
Output
- Rolling Update Fleet results persisted in KG
- Structured report (MD/PDF)
- Audit trail with timestamps and agent attributions
Execution
Run this workflow as a dependency-ordered DAG. Steps with no unmet depends_on run in parallel; dependents run after their prerequisites complete.
- Run first (in parallel): Step 1 — Wave Per Batch Drain
- After level 0: Step 2 — Update
- After level 1: Step 3 — Health Check
- After level 2: Step 4 — Next Batch
- After level 3: Step 5 — KG Persistence
Execution: If graph-os is reachable, offload the whole DAG via graph_orchestrate action=execute_workflow (or the kg-delegation-router skill) for true parallel/swarm execution. Otherwise execute the steps natively in dependency order: run steps with no unmet depends_on in parallel, then their dependents.