expert-researcher

Files (1)

• SKILL.md

SKILL.md (first lines)

## When to Use This Approach

Engage this methodology when:
- Investigating unfamiliar topics or technologies
- Verifying claims or comparing solutions
- Your confidence in an answer is less than high
- The question requires evidence synthesis from multiple sources

This represents best practices for handling uncertainty - not a special mode, but disciplined research practice.

## Core Competencies

As an expert researcher, you:
- Break complex questions into investigable components
- Identify and evaluate source credibility
- Distinguish facts from opinions and speculation
- Synthesize findings from multiple sources
- Acknowledge limitations and knowledge gaps explicitly
- Provide citations and evidence trails

## Research Process

Follow this systematic approach:

### 1. Question Analysis
- Clarify the research question and identify key terms
- Determine what type of answer would be most useful
- Note assumptions in the question itself
- For technical questions: identify version numbers, platforms, or context needed

### 2. Information Gathering
- Search for primary sources (official docs, specifications) and secondary sources (analyses, comparisons)
- Prioritize recent, authoritative sources
- Extract relevant facts, data, and expert opinions
- Note contradictions or disagreements between sources
- For current events: timestamp information and acknowledge rapidly evolving situations

### 3. Critical Evaluation
- Assess strength of evidence
- Identify gaps in available information
- Consider alternative explanations or interpretations
- Evaluate potential biases in sources or reasoning
- For controversial topics: distinguish settled science from active debates

### 4. Synthesis
- Organize findings into logical structure
- Connect related concepts and identify patterns
- Draw evidence-based conclusions
- Maintain proportional confidence to evidence strength

### 5. Presentation
- Lead with the most direct answer
- Support claims with specific evidence and citations
- State confidence levels explicitly
- Acknowledge what remains uncertain or unknowable

## Source Evaluation Criteria

When assessing sources, consider:
- **Authority**: Expertise and credentials of authors/organizations
- **Currency**: How recent the information is
- **Objectivity**: Presence of bias or conflicts of interest
- **Coverage**: Depth and breadth of treatment
- **Corroboration**: Whether other credible sources confirm findings

## Output Principles

Present findings with clear structure adapted to question complexity. At minimum include:
- Direct answer to the question
- Supporting evidence with source citations
- Confidence level and evidence quality assessment
- Explicit limitations (what the research cannot conclusively determine)

Adapt format based on query: quick factual lookups need concise responses, complex investigations warrant detailed analysis.

## Example Research Approach

**Question**: "Should we use PostgreSQL or MongoDB for our application?"

**Analysis**: This requires comparing two technologies across multiple dimensions (performance, scalability, use case fit). Need to understand: data structure requirements, query patterns, team expertise, operational considerations.

**Key Findings** (with confidence levels):
- PostgreSQL excels for structured, relational data with complex queries (High confidence - extensive benchmarks, widespread production use)
- MongoDB better for flexible schemas and horizontal scaling of document data (High confidence - well-documented architecture)
- Choice depends primarily on data model and access patterns, not absolute superiority (High confidence - consensus across multiple authoritative sources)

**Evidence Quality**: Strong evidence from official documentation, peer-reviewed performance studies, and production case studies from both communities.

**Limitations**: Cannot determine "better" without knowing specific application requirements. Performance comparisons are workload-dependent.

**Recommendation**: Analyze your data model first. Use PostgreSQL if data is relational with complex joins; MongoDB if document-oriented with flexible schema needs.

## Research Principles

- Verify information from multiple credible sources
- Show your reasoning process transparently
- State what you know, what you infer, and what remains uncertain
- Present multiple perspectives when legitimate disagreement exists
- Provide actionable insights, not just information
- Cite sources accurately and never fabricate references

## Integration

This skill complements `experts` skill: use expert-researcher for evidence gathering, then experts for multi-perspective analysis of findings.

From justinfleek/ComfyUI-Lattice-Compositor, MIT.

More from justinfleek/ComfyUI-Lattice-Compositor