---

name: research-intelligence description: Extract insights, analyze claims, and synthesize knowledge from research content. Use when processing academic papers, articles, podcasts, videos, meeting transcripts, or any content where the goal is to extract wisdom, analyze arguments, summarize findings, or compile references. Triggers include "analyze this paper", "extract key insights", "summarize the research", "what are the main claims", "extract wisdom from", "compile references", "critique this argument".

Research Intelligence

Transform research content into actionable knowledge using structured extraction patterns.

Pattern Selection

Determine the appropriate pattern based on user intent:

Decision flow:

1. Is it a scientific paper? → analyze_paper (deep) or summarize_paper (quick)
2. Does it contain claims to verify? → analyze_claims
3. Need citations/references? → extract_references
4. General content for insights? → extract_wisdom

Pattern Execution

Each pattern follows the Fabric structure. Load the appropriate reference file, then apply its IDENTITY, STEPS, and OUTPUT format to the user's content.

Core Patterns

See references/ for full pattern definitions:

• analyze_paper.md - Scientific paper deep analysis
• summarize_paper.md - Quick paper summary
• extract_wisdom.md - General insight extraction
• analyze_claims.md - Truth claim evaluation
• extract_references.md - Citation compilation

Output Guidelines

• Use bulleted lists for extracted items (not numbered)
• Never repeat ideas, insights, or quotes
• Don't start consecutive items with the same words
• Match output verbosity to user's apparent needs
• For papers: always note limitations and methodology quality

Chaining Patterns

Offer logical follow-ups after completing a pattern:

• After analyze_paper → offer extract_references for bibliography
• After extract_wisdom → offer to create flashcards or summary
• After analyze_claims → offer to find supporting/contradicting evidence