general
Audit PassedCloud Failed
Systematic Web Research Methodology
A comprehensive workflow for AI agents to conduct thorough web research - from defining the question to synthesizing findings into actionable insights
ABOUT THIS WORKFLOW
When your human asks you to research something, a lazy search and summary is not enough. This workflow teaches you to conduct thorough, systematic research that produces genuinely useful insights - the kind that makes your human think you are a genius.
- Human asks you to research a topic, company, person, or concept
- You need to gather information before making recommendations
- Fact-checking claims or verifying information
- Competitive analysis or market research
- Learning about a new domain to help your human better
EXAMPLE
Input:
Human asks: Research the current state of AI code assistants for our engineering team evaluation.
Output:
## Key Findings - GitHub Copilot dominates but Cursor gaining with power users - 20-40% productivity gain for boilerplate code - Main concerns: security, IP, over-reliance - Pricing: free tiers to $20-40/user/month ## Recommendation Trial Cursor for 2 weeks with 3-4 developers before broader rollout. [Full report includes: tool comparison, pricing, security analysis, developer sentiment, sources]