Agent instructions
AuraScore 81/100

Multi-Agent Handoff and State Synchronization Matrix

Design deterministically governed context passing, shared memory management, and role-boundary enforcement across agent swarms.

Deploy this framework when structuring complex multi-agent workflows where context degradation across handoffs threatens execution integrity. It provides strict schemas for message payloads, role transitions, and conflict mediation.

Template

Role: Distributed AI Systems Lead specializing in multi-agent orchestration, swarm protocols, and context state governance.

Context

  • Multi-agent swarm topology: {{swarm_topology_model}}
  • Roster of specialized agents: {{participating_agent_roster}}
  • Shared memory and context architecture: {{shared_memory_architecture}}
  • Conflict arbitration policy: {{conflict_arbitration_rule}}
  • Maximum workflow timeout window: {{execution_timeout_window}}
  • Telemetry and tracing platform: {{audit_telemetry_standard}}

Task

Design a multi-agent orchestration framework that governs handoff contracts, state serialization, message routing, and execution deadlock prevention across {{participating_agent_roster}} in accordance with {{swarm_topology_model}}.

Method

  1. Define individual agent domain boundaries, capabilities, and explicit non-goals for every persona in {{participating_agent_roster}}.
  2. Map valid state transition paths and handoff triggers across the agent swarm topology.
  3. Establish standard serialized handoff packet schemas containing task context, artifact payloads, confidence scores, and dependency pointers.
  4. Design context delta update procedures for {{shared_memory_architecture}} to avoid race conditions or token memory bloat.
  5. Formalize deadlock detection and circular dependency interruption mechanisms aligned with {{execution_timeout_window}}.
  6. Structure a deterministic conflict resolution protocol implementing {{conflict_arbitration_rule}} for split decisions between peer agents.
  7. Detail event emission specifications compatible with {{audit_telemetry_standard}} for full lifecycle tracing.

Constraints

  • MUST mandate immutable task-id and lineage tracking in every inter-agent message payload.
  • Agents MUST NOT transition workflow control without serializing complete intermediate state artifacts.
  • Shared memory writes MUST require state-lock validation to prevent race conditions in {{shared_memory_architecture}}.
  • Execution chains MUST terminate automatically if unresolved within {{execution_timeout_window}}.

Output format

Generate an actionable orchestration matrix comprising:

  1. Agent Capability and Ownership Matrix (Table mapping Agent, Primary Domain, Input Contract, Output Contract).
  2. Handoff Payload Specification (Standard JSON schema definition for context passing).
  3. State Synchronization and Lock Protocol (Step-by-step technical lifecycle diagram in text format).
  4. Swarm Master Orchestration Directives (Ready-to-use system instructions, 300-500 words).

Self-review

  • Confirm every handoff edge between agents in {{participating_agent_roster}} has an explicit transition trigger.
  • Ensure payload schema accounts for telemetry fields required by {{audit_telemetry_standard}}.
  • Validate that arbitration steps under {{conflict_arbitration_rule}} prevent infinite loop handoffs.
AuraScore breakdown
81/100Provisional
Instruction clarity15/15 · Strong

Explicit role, a named task, and discrete steps the model can follow.

Context architecture12/12 · Strong

Background, inputs and variables the model needs before it starts.

Constraint engineering12/12 · Strong

Hard boundaries — what the model must and must not do.

Output specification6/14 · Thin

A named, field-level shape for the response.

Reasoning structure10/10 · Strong

Ordered work items that force analysis before an answer.

Model compatibility10/10 · Strong

Length and structure that travel across frontier models.

Token efficiency5/10 · Thin

Signal density — instruction weight without padding.

Reusability7/7 · Strong

Documented variables so the scaffold adapts to new inputs.

Robustness3/5 · Adequate

Quality bar, assumptions and behaviour when inputs are thin.

Observed performance1/5 · Thin

How much real usage the template has behind it.

ai-agents
agents-instructions
autonomous-agents-workflows
multi-agent
orchestration
state-management