openclaw-team-builder-skill（OpenClaw team 构建器技能）

Team Builder

Compose the right team for any job by drawing from three rosters of specialists. The Planner analyzes incoming work and proposes an optimal team; the Reviewer validates deliverables before sign-off.

Quick Start

1. Receive a task — any job, project, or request
2. Read PLANNER.md in this skill folder — follow the workflow to classify the domain and propose a team
3. Activate specialists — load the relevant agent definition from the reference files listed below
4. Execute — hand off work through the team using the handoff templates
5. Review — read REVIEWER.md and run the QA workflow before final delivery

The Three Rosters

1. Core Team (TEAM-CORE.md)

The permanent OpenClaw agents. Always available, always running.

2. Agency Division (TEAM-AGENCY.md)

55+ specialist agents across 9 divisions. Activated on demand from reference/agency-agents-main/.

3. Research Lab (TEAM-RESEARCH.md)

Autonomous experiment methodology adapted from Karpathy's autoresearch. Run iterative experiment loops on any measurable problem.

Applicable to: trading strategy optimization, image analysis pipelines, model tuning, data analysis, any domain with a measurable metric.

Cross-Team Workflows

The real power is mixing specialists across rosters. Here are proven combinations:

Trading Strategy Optimization

IG (market context) + Research Lab (experiment loop) + AI Engineer (model tuning)
→ IG provides live market data and strategy parameters
→ Research Lab runs 5-min experiment iterations on backtests
→ AI Engineer tunes neural network parameters
→ Reviewer validates with evidence before deploying to live

Visual Content Pipeline

Artist (image generation) + Image Prompt Engineer (prompt crafting) + Visual Storyteller (narrative)
→ Image Prompt Engineer crafts detailed, structured prompts
→ Artist generates via xAI grok-imagine-image-pro
→ Visual Storyteller evaluates narrative quality
→ Iterate until quality threshold met

Astronomy / Image Analysis

Artist (image capture/generation) + Research Lab (analysis loop) + AI Engineer (classification)
→ Artist handles image acquisition and enhancement
→ Research Lab runs iterative analysis (feature detection, classification)
→ AI Engineer builds/tunes detection models
→ Results feed back for next iteration

Dashboard / UI Feature Build

Senior PM (scope) + Frontend Developer (build) + Evidence Collector (QA) + Reality Checker (sign-off)
→ PM breaks spec into tasks with acceptance criteria
→ Frontend Developer implements mobile-first
→ Evidence Collector screenshots and validates each task
→ Reality Checker gives final production-readiness verdict

Full Product Launch

CEO (orchestrate) + Engineering (build) + Design (UX) + Marketing (launch) + Testing (validate)
→ CEO activates Planner to scope the project
→ Engineering + Design work in parallel (Dev↔QA loops)
→ Marketing prepares launch materials
→ Reviewer signs off before go-live

Activating a Specialist

To activate any Agency specialist, load their definition file:

Read the agent definition at:
reference/agency-agents-main/[division]/[agent-file].md

Then adopt that agent's:
- Identity and personality
- Core mission and rules
- Workflow process
- Success metrics

File paths follow the pattern:

• reference/agency-agents-main/engineering/engineering-frontend-developer.md
• reference/agency-agents-main/design/design-image-prompt-engineer.md
• reference/agency-agents-main/testing/testing-evidence-collector.md

See TEAM-AGENCY.md for the complete index with all file paths.

Handoff Protocol

When passing work between specialists, use this template:

## Handoff
| Field | Value |
|-------|-------|
| From | [Agent Name] |
| To | [Agent Name] |
| Task | [What needs to be done] |
| Priority | [Critical / High / Medium / Low] |

## Context
- Current state: [What's been done]
- Relevant files: [File paths]
- Constraints: [Limits, requirements]

## Deliverable
- What is needed: [Specific output]
- Acceptance criteria:
  - [ ] [Criterion 1]
  - [ ] [Criterion 2]

## Quality
- Evidence required: [What proof of completion looks like]
- Reviewer: [Who validates this deliverable]

For complete handoff templates (QA pass/fail, escalation, phase gates), see: reference/agency-agents-main/strategy/coordination/handoff-templates.md

NEXUS Pipeline Modes

For larger projects, use the NEXUS pipeline from the Agency framework:

Pipeline phases: Discover → Strategize → Scaffold → Build → Harden → Launch → Operate

Quality gates between every phase. Evidence required for all assessments.

For full NEXUS strategy: reference/agency-agents-main/strategy/nexus-strategy.md For activation prompts: reference/agency-agents-main/strategy/coordination/agent-activation-prompts.md For quickstart: reference/agency-agents-main/strategy/QUICKSTART.md

Reference Files in This Skill