GPT-4o
⚙️ Technical
Advanced
Orchestrate a Multi-Agent Research Pipeline
Design a coordinator agent that spawns specialized sub-agents for [RESEARCH_DOMAIN], aggregates findings, and produces a unified structured report.
The Prompt
# Orchestrate a Multi-Agent Research Pipeline You are a multi-agent systems architect. Design a coordinator that orchestrates a team of specialized sub-agents to research [RESEARCH_DOMAIN]. ## Coordinator Role The coordinator accepts "[RESEARCH_QUESTION]" and breaks it into [SUB_TASK_COUNT] parallel sub-tasks, one per specialist agent. ## Sub-Agent Definitions Define each specialist: - **[AGENT_1_NAME]**: scope [SCOPE_1], tools: [TOOL_LIST_1] - **[AGENT_2_NAME]**: scope [SCOPE_2], tools: [TOOL_LIST_2] - **[AGENT_3_NAME]**: scope [SCOPE_3], tools: [TOOL_LIST_3] ## Execution Plan 1. Coordinator dispatches sub-tasks in parallel via [ORCHESTRATION_FRAMEWORK] 2. Each agent reports results as structured JSON to [RESULTS_STORE] 3. Coordinator polls until all agents complete or [TIMEOUT] is reached 4. Coordinator synthesizes all findings into a unified report ## Conflict Resolution When agents return conflicting data, apply rule: [CONFLICT_RESOLUTION_RULE]. ## Output Deliver the final report to [RECIPIENT] in [FORMAT] with per-section source citations.
📝 Fill in the blanks
Replace these placeholders with your own content:
[RESEARCH_DOMAIN]
[RESEARCH_QUESTION]
[SUB_TASK_COUNT]
[AGENT_1_NAME]
[SCOPE_1]
[TOOL_LIST_1]
[AGENT_2_NAME]
[SCOPE_2]
[TOOL_LIST_2]
[AGENT_3_NAME]
[SCOPE_3]
[TOOL_LIST_3]
[ORCHESTRATION_FRAMEWORK]
[RESULTS_STORE]
[TIMEOUT]
[CONFLICT_RESOLUTION_RULE]
[RECIPIENT]
[FORMAT]
How to use this prompt
1
Copy the prompt
Click "Copy Prompt" above to copy the full prompt text to your clipboard.
2
Replace the placeholders
Swap out anything in [BRACKETS] with your specific details.
3
Paste into GPT-4o
Open your preferred AI assistant and paste the prompt to get started.