Multi-Agent Workflow State Transition Matrix

Role: Principal Workflow Automation Engineer specializing in deterministic orchestration, distributed agent coordination, and state machine design.

Context

• Business Domain: {{workflow_domain}}
• Agent Ecosystem: {{primary_agent_cluster}}
• Orchestration Framework: {{orchestration_engine}}
• Performance Ceiling: {{max_latency_budget}}
• Integrated Services: {{external_api_dependencies}}
• Observability Target: {{telemetry_sink}}

Task

Generate a Multi-Agent Workflow State Transition Matrix for {{workflow_domain}} using {{orchestration_engine}} to establish strict stage gate transitions, agent-to-agent handoffs, data contract transfers, and fallback routines within {{max_latency_budget}}.

Method

1. Break down the {{workflow_domain}} workflow into sequential and parallel execution states.
2. Assign responsible nodes from {{primary_agent_cluster}} to each distinct state transition.
3. Define input schema requirements, state payloads, and output validation rules for each transition boundary.
4. Account for network latency and rate limits imposed by {{external_api_dependencies}}.
5. Establish retry limits, backoff strategies, and dead-letter queue routing for each transition.
6. Formulate deterministic state exit conditions ensuring performance stays within {{max_latency_budget}}.
7. Map monitoring event hooks for continuous ingestion into {{telemetry_sink}}.

Constraints

• State handoffs MUST define explicit data payload keys and validation constraints.
• The workflow MUST NOT allow circular agent handoff loops without strict recursion ceilings.
• Include both successful transition criteria and failed transition degradation paths.
• Limit matrix entries to critical lifecycle transitions within the pipeline.

Output format

Structure the deliverable into two parts:

1. Architecture Overview: A concise 3-paragraph summary covering orchestration logic, state persistence, and timeout policies.
2. State Transition Matrix: A markdown table containing 6 columns (Current State, Delegating Agent, Target State & Agent, Entry/Exit Conditions, Latency Budget, Fallback/Retry Action) covering 5-8 sequential state steps.

Self-review

• Ensure total accumulated latency across states fits within {{max_latency_budget}}.
• Confirm that every agent from {{primary_agent_cluster}} has an assigned role.
• Validate that all fallback actions have defined target recovery states.