Autonomous Research Pipeline for Organizations

# Autonomous Research Pipeline for Organizations

You are a research systems architect building a fully autonomous research pipeline for [RESEARCH_ORGANIZATION] studying [RESEARCH_DOMAIN].

## Stage 1 — Question Generation Agent

Write the system prompt for an agent that identifies the highest-priority research questions in [RESEARCH_DOMAIN] each cycle. Inputs: recent news context, last cycle's findings, standing research objectives. Output: ranked list of 5 questions with rationale.

## Stage 2 — Source Discovery Agent

Write the discovery prompt that finds relevant sources for each research question:
- Search strategy and query templates for [RESEARCH_DOMAIN]
- Quality scoring (relevance, recency, authority, novelty)
- Deduplication and source diversity rules

## Stage 3 — Reading and Extraction Agent

Write extraction prompts for different source types:
- Academic papers: extract claims, methodology, key findings, limitations
- News articles: extract facts, named entities, timeline, source credibility
- Primary sources: extract verbatim quotes and context

Define the JSON output schema for [RESEARCH_DOMAIN].

## Stage 4 — Synthesis Agent

Write the synthesis prompt that combines multiple sources:
- How to handle contradictory findings
- Confidence attribution per claim
- Output format: structured research memo with citations

## Stage 5 — Gap Analysis Agent

Write the agent that identifies what's still unknown after synthesis and generates follow-up questions for the next cycle.

## Orchestration Design

Define how the 5 agents hand off to each other in [TOOL_PLATFORM], with failure handling and a delivery schedule for [TEAM_NAME].