APromptAura

#multi-agent

Top prompts tagged multi-agent.

0
p/agentsAauraNovice4 days agohas example⁂auto-checked

Handoff protocol between specialized agents with no context loss

Design a handoff protocol for when one specialized agent (agent_a_role) needs to pass a task to another (agent_b_role) mid-conversation. Requirements: 1. What exactly gets handed off: define the minimum sufficient package (not the full conversation dump) so agent B can pick up correctly, including what agent A already tried/ruled out so agent B doesn't repeat wasted work. 2. The trigger condition for handoff: specific signals that this is now outside agent A's competence (not a vague 'if it seems hard'), tailored to what agent A vs agent B are each actually good at. 3. What agent B is required to acknowledge before proceeding: confirm it understood the handed-off context correctly, in a form the user can also see, so a miscommunication is caught before agent B acts on a misunderstanding. 4. How the user experiences this: should the handoff be visible/announced to the user or seamless, and why, given the stakes of this specific handoff. 5. What happens if agent B determines the handoff was wrong (task actually still belongs with agent A, or belongs with a third agent) - avoid a ping-pong loop. Agent A role: agent_a_role Agent B role: agent_b_role Domain: domain
ClaudeGPT#multi-agent#handoff#agent-design
0
0
p/agentsAauraNovice4 days agohas example💀failed our automated check

Manager/worker multi-agent orchestration design for a complex task

Design a multi-agent system (manager + worker agents) for the task: complex_task_description. I need the actual role split and communication protocol, not just 'have a coordinator agent'. Cover: 1. What the manager agent is responsible for versus what it delegates, with a rule for how it decides to split the task into subtasks (by data partition, by pipeline stage, by specialization) given this specific task. 2. The exact message format passed from manager to worker (what context each worker needs to do its job without the full conversation history) and from worker back to manager (result plus confidence/caveats, not just raw output). 3. How the manager verifies a worker's output before using it or passing it downstream, specific to this task's likely failure modes, not a generic 'sanity check'. 4. What happens when a worker fails or returns a low-confidence result: retry with a different worker, retry with more context, or escalate to the manager's own reasoning versus to a human. 5. How to prevent workers from duplicating work or working on stale/conflicting versions of shared state, if multiple workers run in parallel. Task: complex_task_description Constraints (latency, cost, number of parallel workers): constraints
ClaudeGPT#multi-agent#orchestration#system-design
0