# Summon Team Research

> Launch a research team (Fetcher + BrainStorm + specialist) for deep investigation

- Skill: `majiayu000/summon-team-research` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add majiayu000/summon-team-research`
- Raw SKILL.md: https://api.skillmd.com/api/skills/majiayu000/summon-team-research/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Research & Search
- Author: majiayu000 (https://skillmd.com/u/majiayu000)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/majiayu000/summon-team-research

---


# /summon-team-research - Research Team

Launch a coordinated research team: Fetcher gathers sources, BrainStorm generates ideas, and a specialist agent provides domain expertise — working in parallel.

## Usage

```
/summon-team-research <topic> [specialist=<agent>]
```

If no specialist is specified, Orca (you, the Team Lead) fills the synthesis role.

## Examples

```
/summon-team-research WebSocket vs SSE for real-time updates
/summon-team-research Hebrew NLP libraries specialist=builder
/summon-team-research EU AI Act compliance requirements specialist=legal
/summon-team-research Competitor analysis of Cursor IDE specialist=gal
```

## Instructions

When invoked, you are the **Team Lead** (Orca). Spawn teammates to research the given topic.

### Step 1: Parse the Request

- Extract the research topic
- Identify the specialist agent (default: none, you synthesize)
- Check `Library/Sources/` for existing research on this topic

### Step 2: Spawn Teammates

Launch teammates using the Task tool.

#### Teammate 1: Fetcher

```
You are Fetcher — The Researcher.

READ FIRST:
- Your identity: Library/Fetcher/Fetcher_Identity.md
- Your knowledge: Library/Knowledge/Fetcher/README.md
- CLAUDE.md is auto-loaded (shared protocols)

YOUR TASK:
Research: [TOPIC]

DELIVERABLE:
1. Save source files to Library/Sources/[topic-slug]/
   - One .md file per source (clean, attributed, dated)
   - README.md index of all sources gathered
2. Create a study file at Dashboard/Work_Space/[Topic]_Study.md
   - Executive summary
   - Key findings organized by theme
   - Source attribution for every claim

GUIDELINES:
- Use WebSearch and WebFetch to gather information
- Respect robots.txt and rate limits
- Prefer official documentation, reputable sources
- Minimum 3 sources, aim for 5-8
- After delivering, update Library/Knowledge/[requesting-agent]/README.md with links to new sources

COORDINATION:
- BrainStorm will use your research to generate ideas
- [Specialist] will provide domain-specific analysis
- Focus on FACTS and SOURCES — leave the creative thinking to BrainStorm
```

#### Teammate 2: BrainStorm

```
You are BrainStorm — The Creative.

READ FIRST:
- Your identity: AgenTeam/BrainStorm/BrainStorm_Identity.md
- Your knowledge: Library/Knowledge/BrainStorm/README.md
- CLAUDE.md is auto-loaded (shared protocols)

YOUR TASK:
Generate ideas and insights about: [TOPIC]

Start by reading any existing sources in Library/Sources/ related to this topic. When Fetcher's research arrives at Dashboard/Work_Space/[Topic]_Study.md, incorporate those findings.

DELIVERABLE:
Write to Dashboard/Work_Space/BRAINSTORM_[topic].md

Include:
- At least 5 distinct ideas or approaches
- Pros/cons for each
- "What if..." explorations
- Connections to AutoMates and our current projects
- Recommended direction with reasoning

COORDINATION:
- Fetcher is gathering the facts — build on them
- [Specialist] will evaluate from their domain
- Be bold — this is the place for creative thinking
```

#### Teammate 3: Specialist (if specified)

```
You are [Specialist Agent] — [Role].

READ FIRST:
- Your identity: [identity path]
- Your knowledge: Library/Knowledge/[Agent]/README.md
- CLAUDE.md is auto-loaded (shared protocols)

YOUR TASK:
Provide [domain] analysis of: [TOPIC]

Read Fetcher's research when available at Dashboard/Work_Space/[Topic]_Study.md.
Read BrainStorm's ideas when available at Dashboard/Work_Space/BRAINSTORM_[topic].md.

DELIVERABLE:
Write to Dashboard/Work_Space/[AGENT]_Analysis_[topic].md

Include:
- Domain-specific assessment of the research findings
- Feasibility analysis of BrainStorm's ideas
- Risks and concerns from your [domain] perspective
- Recommendations

COORDINATION:
- Fetcher gathered the sources
- BrainStorm generated creative ideas
- Your job is the expert evaluation from a [domain] lens
```

### Step 3: Synthesize

When all teammates finish, create a synthesis:

```
=== RESEARCH TEAM COMPLETE ===

TOPIC: [topic]

SOURCES: [count] sources gathered by Fetcher
  Location: Library/Sources/[topic-slug]/

IDEAS: [count] ideas generated by BrainStorm
  Location: Dashboard/Work_Space/BRAINSTORM_[topic].md

SPECIALIST: [Agent]'s assessment
  Location: Dashboard/Work_Space/[Agent]_Analysis_[topic].md

KEY FINDINGS:
1. [finding 1]
2. [finding 2]
3. [finding 3]

RECOMMENDED DIRECTION:
[synthesis of research + ideas + specialist input]

NEXT STEPS:
- [what to do with this research]
===============================
```

## Notes

- Fetcher starts immediately (no dependencies)
- BrainStorm can start independently but should incorporate Fetcher's work when available
- Specialist waits for both Fetcher and BrainStorm when possible
- All deliverables go to `Dashboard/Work_Space/` for team visibility
- Library/Sources/ gets the permanent research files
- The user decides what to do with the findings (Pilot-in-Command)