Matrix Deep Research

Perform comprehensive research on a topic using multiple sources: web search, Context7 documentation, Matrix memory, GitHub repositories, and more.

Usage

Parse the arguments: $ARGUMENTS

Format: <query> [depth]

• query: The research topic or question
• depth (optional): quick | standard | exhaustive (default: standard)

Depth Levels

• quick: Fast research using 2-3 sources, ~500 words output
• standard: Balanced research using 4-5 sources, ~1500 words output
• exhaustive: Deep research using 6+ sources, ~3000+ words output

Research Pipeline

Phase 1: Query Expansion

Analyze the query to:

1. Identify the core topic and intent
2. Generate 2-4 related sub-queries for comprehensive coverage
3. Identify relevant technical domains (frameworks, languages, tools)

Phase 2: Multi-Source Gathering

Collect information from multiple sources in parallel:

1. Web Search (WebSearch tool)

   • Search for recent articles, tutorials, and discussions
   • Focus on authoritative sources (official docs, reputable blogs)

2. Context7 Documentation (mcp__plugin_matrix_context7__resolve-library-id + mcp__plugin_matrix_context7__query-docs)

   • If the topic involves a library/framework, query Context7
   • Get up-to-date API documentation and examples

3. Matrix Memory (matrix_recall)

   • Search for relevant past solutions and patterns
   • Include related failures to avoid

4. GitHub Repositories (matrix_repomix)

   • For implementation patterns, fetch relevant code from authoritative repos
   • Use Phase 1 to get file suggestions, confirm with user before packing

5. Additional Sources (for exhaustive depth)

   • GitHub Issues/Discussions via gh CLI
   • Stack Overflow via web search

Phase 3: Content Fetching

For promising URLs from web search:

1. Use WebFetch to retrieve full content
2. Extract key information and code examples
3. Note the source for citations

Phase 4: Synthesis

Combine all gathered information:

1. Deduplicate overlapping information
2. Organize by subtopic/theme
3. Identify consensus vs. conflicting approaches
4. Highlight best practices and anti-patterns
5. Include relevant code examples

Phase 5: Output

Generate a polished markdown document:

# Research: [Topic]

> Generated by Matrix Deep Research on [date]
> Depth: [quick|standard|exhaustive]

## Summary
[2-3 sentence overview]

## Key Findings
### [Subtopic 1]
[Content with code examples]

### [Subtopic 2]
[Content with code examples]

## Best Practices
- [Practice 1]
- [Practice 2]

## Common Pitfalls
- [Pitfall 1]
- [Pitfall 2]

## Code Examples
[Consolidated, working examples]

## Sources
1. [Source 1](url)
2. [Source 2](url)
...

Save to session directory: Use $CLAUDE_SESSION_DIR/matrix-research-[slug]-[timestamp].md

If $CLAUDE_SESSION_DIR is not available, fall back to current working directory: ./matrix-research-[slug]-[timestamp].md

Examples

/matrix:deep-research React Server Components best practices standard
/matrix:deep-research "how to implement OAuth 2.0 with PKCE" exhaustive
/matrix:deep-research TypeScript generics quick

Output

After completing research:

1. Display a summary of findings in the chat
2. Report the full markdown file location
3. Offer to elaborate on any section