Execute complete 7-phase deep research workflow by delegating to the research-orchestrator-agent. Thin wrapper skill that ensures proper agent invocation with structured research prompts.
The Research Executor is a thin wrapper skill that delegates research execution to the research-orchestrator-agent. It validates inputs, prepares the execution context, and invokes the autonomous orchestrator agent to handle the complete 7-phase deep research workflow.
User Request
↓
research-executor skill (this skill - thin wrapper)
↓
research-orchestrator-agent (autonomous agent)
↓
├── Phase 1: Question Refinement
├── Phase 2: Research Planning
├── Phase 3: Multi-Agent Deployment
├── Phase 4: Source Triangulation
├── Phase 5: Knowledge Synthesis
├── Phase 6: Quality Assurance
└── Phase 7: Output Generation
Key Change: All orchestration logic has been moved to research-orchestrator-agent. This skill only handles:
Execute research using structured prompt:
[STRUCTURED_PROMPT]
The executor will:
1. Validate prompt structure
2. Invoke research-orchestrator-agent
3. Monitor progress
4. Return results from RESEARCH/[topic]/
Required: Structured research prompt with:
Optional:
RESEARCH/[topic]/
├── README.md
├── executive_summary.md
├── full_report.md
├── data/
│ ├── statistics.md
│ └── ontology/
├── sources/
│ ├── bibliography.md
│ └── source_quality_table.md
├── research_notes/
│ └── agent_findings_summary.md
└── appendices/
├── methodology.md
└── limitations.md
| Error Code | Description | Action |
|---|---|---|
| E001 | Incomplete structured prompt | Request missing fields |
| E002 | Agent deployment failed | Retry with fallback config |
| E003 | Agent execution timeout | Report partial results |
| E004 | Quality threshold not met | Trigger refinement (max 2 attempts) |
| Limit | Value | Enforced By |
|---|---|---|
| Max parallel agents | 8 | research-orchestrator-agent |
| Max research time | 90 minutes | research-orchestrator-agent |
| Min quality score | 8.0 | research-orchestrator-agent |
| Max token per agent | 15,000 | research-orchestrator-agent |
Primary Agent: research-orchestrator-agent
Supporting Agents (invoked by orchestrator):
got-agent: For complex research optimizationsynthesizer-agent: For findings aggregationred-team-agent: For quality validationontology-scout-agent: For domain reconnaissanceSee examples.md for usage scenarios.
See instructions.md for implementation guide.