Deep Research Command

You are initiating a deep research workflow using the orchestrator-workers pattern.

IMPORTANT: Load the research-orchestrator skill first, then spawn the research-orchestrator agent to handle the full workflow.

Workflow

1. Load Skill: Use the Skill tool to load "research-orchestrator:research-orchestrator" for context
2. Spawn Orchestrator: Use the Task tool with subagent_type="research-orchestrator" to delegate the research

The orchestrator agent will:

• Ask clarifying questions about scope and requirements
• Identify count requirements (e.g., "10 competitors")
• Delegate to schema-builder to create research structure
• Spawn parallel researcher subagents
• Run validation and fix issues
• Assemble final output

Example Invocations

• /research AI code assistant market competitive analysis with 10 competitors
• /research Market research on sustainable packaging industry
• /research Technical deep-dive on WebAssembly performance vs JavaScript

Output Location

Research output will be saved to:

• Schemas: output/schemas/
• Sections: output/sections/
• Validation: output/validation/
• Final: output/final/report.html and output/final/report.json

Spawn the orchestrator now with the user's research topic.