Research Coordinator Partner

An agent specialized in strategic research planning and multi-researcher orchestration, breaking down complex research requirements into optimally distributed tasks across specialist researchers with proper sequencing and quality control.

When to Use This Agent

Choose Research Coordinator when:

• Managing multiple research agents working on a complex topic
• Distributing research tasks based on specialization and expertise
• Sequencing research phases to optimize quality and efficiency
• Consolidating findings from parallel research streams
• Resolving conflicts and gaps between researcher outputs

Consider alternatives when:

• Creating research briefs from queries (use a brief generator agent)
• Conducting individual research tasks (use a research analyst agent)
• Synthesizing findings into reports (use a report generator agent)

Quick Start

Copy
# .claude/agents/research-coordinator.yml
name: Research Coordinator
model: claude-sonnet-4-20250514
tools:
  - Read
  - Write
  - Bash
  - Glob
  - Grep
prompt: |
  You are a research coordinator. Plan and orchestrate multi-researcher
  investigations. Assign tasks based on expertise, manage dependencies,
  ensure quality, and consolidate findings. Balance thoroughness with
  efficiency. Identify gaps and conflicts in research outputs.

Example invocation:

Copy
claude --agent research-coordinator "Coordinate a research team
  investigating the feasibility of migrating from monolith to
  microservices. Assign technical, organizational, and financial
  research streams. Manage dependencies and consolidate findings."

Core Concepts

Coordination Workflow

Research Brief → Task Decomposition → Assignment → Monitoring → Consolidation
      │                │                  │            │              │
  Objectives      Sub-tasks          Match to      Track         Merge
  Constraints     Dependencies       expertise     progress      findings
  Timeline        Sequencing         Allocate      Quality       Resolve
                                     resources     check         conflicts

Task Distribution Matrix

Factor	How It Affects Assignment
Domain expertise	Match researcher to their specialty area
Task complexity	Allocate more time for complex investigations
Dependencies	Sequence dependent tasks, parallelize independent ones
Source access	Assign based on available databases and contacts
Time sensitivity	Assign urgent tasks to available researchers
Quality requirements	Assign critical findings to senior researchers

Coordination Patterns

Pattern 1: Parallel Investigation
  Research A ──→ ┐
  Research B ──→ ├── Consolidate → Report
  Research C ──→ ┘

Pattern 2: Sequential Deepening
  Broad Survey ──→ Gap Analysis ──→ Deep Dives ──→ Report

Pattern 3: Iterative Refinement
  Research ──→ Review ──→ Refine ──→ Research ──→ Review ──→ Report

Pattern 4: Expert Validation
  Research ──→ Draft Findings ──→ Expert Review ──→ Revised Findings

Configuration

Parameter	Description	Default
max_parallel_tasks	Maximum concurrent research streams	3
coordination_style	Orchestration pattern	Parallel + consolidate
quality_checkpoints	Review points during research	After each phase
conflict_resolution	How to handle contradicting findings	Document and analyze
gap_detection	Actively identify missing coverage	true
status_frequency	Progress update interval	Per-task completion
escalation_policy	When to escalate issues	Blockers and conflicts

Best Practices

1. Create mutually exclusive, collectively exhaustive task assignments. Each research task should cover a distinct area with no overlaps (mutually exclusive) and the full set of tasks should cover the entire research scope (collectively exhaustive). Overlapping assignments create redundant work and conflicting findings. Missing coverage creates gaps that aren't discovered until consolidation.

2. Sequence tasks by information dependency, not by difficulty. If the financial analysis needs the technical feasibility assessment as input, schedule technical research first. If organizational readiness research is independent of technical research, run them in parallel. Map dependencies explicitly and schedule the critical path through the most dependent chain.

3. Build quality checkpoints between research phases. Don't wait until all research is complete to discover quality issues. Review findings after each major phase: are sources credible? Are claims supported? Are there gaps in coverage? Early quality checks prevent wasted effort from low-quality research that would need to be redone anyway.

4. Track and surface conflicts between researchers proactively. When Researcher A finds "migration takes 6-12 months" and Researcher B finds "migration takes 2-3 years," this isn't a mistake—it's valuable information about different contexts, assumptions, or methodologies. The Coordinator should identify such conflicts, analyze their causes, and present the reconciled view rather than leaving contradictions for the final reader.

5. Maintain a shared context document that all researchers access. As research progresses, findings from one stream inform other streams. Create a living document summarizing key findings, assumptions, and emerging themes that all researchers reference. This shared context prevents researchers from working in isolation and discovering late that their assumptions were invalidated by another stream's findings.

Common Issues

Research streams produce inconsistent levels of detail. Set explicit deliverable standards in the task assignment: minimum number of sources, required sections in findings documents, confidence ratings for each claim. Without standards, some researchers produce three-page reports while others produce three-paragraph summaries, making consolidation difficult.

Coordination overhead exceeds the time saved by parallelization. This happens when tasks are too small or too interdependent for parallel execution. The sweet spot is 3-5 parallel research streams, each taking several hours to complete. If tasks can be done in 30 minutes each, sequential execution with a single researcher is more efficient than coordinating three researchers for 30 minutes of work each.

Consolidation discovers fundamental gaps in research coverage. Prevent this by running a coverage check before research begins: map every sub-question in the research brief to a specific research task and researcher. After the first round of findings, run a gap analysis: which questions remain unanswered? Which areas have insufficient evidence? Assign targeted follow-up research for gaps rather than repeating the full research cycle.