Knuckles-Team Systems Issue Troubleshooter Workflow
CONCEPT:INFRA-001
Inspect logs, find zombie/hung processes, disk space shortages, or corruptions, and remediate using Systems Manager
Steps
Step 0: collect-system-health [skill: systems-manager]
Agent: discovery-agent
Tools: tun_tm_system, tun_tm_hosts
Capture general operating system performance stats and health checks using sm_system_operations with action='system_health_check' and action='get_os_statistics'.
Step 1: inspect-processes [skill: systems-manager]
Agent: deployer-agent
Tools: pt_stack, cnt_cm_compose_operations
Retrieve the active process table using sm_process_operations with action='list_processes' to scan for zombie states, memory leaks, or hung process IDs (PIDs).
Step 2: inspect-disk-and-logs [skill: systems-manager]
Agent: verifier-agent
Tools: pt_docker, cnt_cm_container_operations
Retrieve disk partition layouts and usage indicators using sm_disk_operations with action='get_disk_space_report'. Fetch the latest system journal log records using sm_file_operations with action='get_system_logs' and lines=150 to detect underlying filesystem or service corruptions.
Step 3: confirm-remediation [skill: user-interaction]
Agent: dns-configurator
Tools: adg_rewrites, td_zones
Present the diagnostic dashboard (including lists of zombie/hung processes, disk bottlenecks, and log error summaries) to the user. Request choice of corrective intervention (e.g. killing a process, running disk cleanups).
Step 4: apply-approved-remediation [skill: systems-manager]
Agent: discovery-agent
Tools: tun_tm_system, tun_tm_hosts
Execute the approved cleanup commands. Terminate rogue process instances using sm_process_operations with action='kill_process' and target pid. Reclaim storage space using sm_system_operations with action='clean_temp_files' and action='clean_package_cache'.
Step 5: report-remediation [skill: user-interaction]
Agent: deployer-agent
Tools: pt_stack, cnt_cm_compose_operations
Present the revised health diagnostic metrics to verify successful remediation and system stabilization.
Step 6: KG Persistence [depends_on: Step 3, Step 5]
Agent: dns-configurator
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
- Systems Issue Troubleshooter 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 0 — collect-system-health; Step 1 — inspect-processes; Step 2 — inspect-disk-and-logs; Step 3 — confirm-remediation; Step 4 — apply-approved-remediation; Step 5 — report-remediation
- After level 0: Step 6 — KG Persistence
Execution: If graph-os is reachable, offload the whole DAG via graph_orchestrate action=execute_workflow (or the kg-delegate 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.