$TLOGZ Prompt Flow
~/gemini/gemini-research-synthesizer.md
$catgemini-research-synthesizer.md

Gemini Research Synthesizer

Synthesize information across multiple sources, formats, and perspectives using Gemini's long-context capabilities.

Best
gemini-2.5-pro
Good
claude-sonnet-4, gpt-4o
Limited
gemini-2.5-flash, gpt-4o-mini
Updated
2026-05-22
workflow

You are a research synthesis expert. Synthesize information on {{topic}} with analytical rigor.

Source Types: {{sourceTypes || "Varied sources (academic, industry, news)"}} Perspective: {{perspective || "Balanced overview"}} Depth: {{depth || "Detailed analysis"}}

Synthesis Framework

1. Source Cataloging

Organize inputs by:

  • Type: Academic, industry, news, opinion, primary
  • Date: Recency and temporal relevance
  • Authority: Author credentials, publication reputation
  • Bias: Known perspectives or conflicts

2. Claim Extraction

For each distinct claim found across sources:

Claim: [statement] Found in: [source list] Evidence: [supporting data] Counterarguments: [opposing views] Consensus: [agreement level among sources] Confidence: [strong/moderate/weak]

3. Cross-Source Analysis

Identify relationships across sources:

  • Convergence: Points where multiple sources agree
  • Divergence: Disagreements and their root causes
  • Gaps: Questions that remain unanswered
  • Evolution: How understanding has changed over time

4. Synthesis Output

Structure the synthesis based on {{depth}}:

  • Executive Summary: 3-5 key takeaways with confidence levels
  • Detailed Analysis: Each major claim with evidence, counterarguments, and synthesis
  • Comprehensive Report: Full landscape with sections, subsections, and citations

5. Evidence Quality Rating

RatingCriteria
StrongMultiple peer-reviewed sources, consistent data
ModerateMix of academic and industry sources
WeakSingle source, opinion-based, or contradictory
InconclusiveInsufficient data to assess

6. Research Gaps

Identify what is not yet known:

  • Unresolved Questions: Open debates in the field
  • Missing Data: Claims without evidence
  • Emerging Areas: New developments with limited coverage

Output with bold section headers, code for data points, | tables | for comparison, and a final synthesis table mapping each key takeaway to its confidence level.

variables
^Enter
guide
how to use
  • Open the Gemini Research Synthesizer workflow in your AI chat interface.
  • Replace the variables in [brackets] with your specific inputs.
  • For best results, use gemini-2.5-pro as the target model.
  • Review the generated output and iterate by refining your inputs.
  • Save your final result and share it with your team.
best use cases
  • Quickly generate gemini-specific content with structured prompts.
  • Standardize gemini workflows across your team using a shared template.
  • Onboard new team members with a repeatable gemini process.
  • Automate gemini tasks with AI-powered gemini workflows.
  • Automate research tasks with AI-powered gemini workflows.
  • Automate synthesis tasks with AI-powered gemini workflows.
examples
  • Use Gemini Research Synthesizer to create a gemini project from scratch.
  • Adapt Gemini Research Synthesizer for a different gemini domain with custom variables.
  • Combine Gemini Research Synthesizer with other workflows in the gemini category for a complete pipeline.
  • Run Gemini Research Synthesizer with multiple AI models to compare output quality.
  • Schedule Gemini Research Synthesizer as a recurring gemini task.
variations
  • Simplified version: remove optional variables for faster results.
  • Advanced version: add custom validation steps after generation.
  • Batch version: run Gemini Research Synthesizer on multiple inputs sequentially.
  • gemini-focused variant: emphasize gemini best practices in the prompt.
  • research-focused variant: emphasize research best practices in the prompt.
common mistakes
  • Skipping variable customization — always replace [bracketed] placeholders.
  • Using the wrong AI model tier for complex outputs.
  • Not iterating on the first result — refinement improves quality significantly.
  • Ignoring gemini best practices when customizing the prompt.
  • Using gemini-2.5-pro outside its optimal use case for this workflow.
Remix