Agent Team Skill

Manage AI agent team members with skills, roles, and task delegation.

Team Members

List all team member information:

CODEBLOCK0

Common scenarios:

• - Check who is the current team leader
• Find members with specific expertise before task assignment
• Review team structure and available roles

Output example:
CODEBLOCK1

⚡ Task Delegation Rules (Core Principle)

Delegation Timing:

1. 1. First complete prep work: understand requirements, clarify goals, confirm constraints
2. When entering implementation: identify the best person for execution, delegate to them
3. Follow up after delegation: check output quality, ensure requirements are met

Delegation Context (what to pass):
When delegating, always provide:

• - Original requirements and success criteria
• Relevant background and context
• Your execution plan and any constraints
• Expected output format

Delegation Failover:
If teammate fails to complete:

1. 1. First attempt: Send back with specific feedback for revision
2. Second attempt: Reassign to another teammate with adjusted context
3. Third attempt: Escalate to leader OR execute yourself

🔄 Task Processing Flow (Highest Priority)

Plan → Do → Check → Act

IMPORTANT: This is a continuous improvement cycle. If task is incomplete in Act phase, loop back to Plan.

1. Plan — Planning Phase

Goal: Prepare thoroughly, avoid blind execution

• - Understand requirements and clarify questions
• Define goals and success criteria
• Identify risks and determine ownership
• Create execution plan

2. Do — Execution Phase

Goal: Execute the plan while maintaining progress

• - Execute or delegate based on ownership
• Track progress and key decisions

3. Check — Checking Phase

Goal: Verify results against requirements

• - Verify completeness and quality
• Check compliance with standards

4. Act — Acting Phase

Goal: Summarize and decide next steps

• - ✅ Task complete → End
• ❌ Task incomplete → Loop back to Plan

Add/Update Member

Add a new member or update existing member information:

CODEBLOCK2

Parameters:

• - --agent-id: Member unique identifier (required)
• INLINECODE1: Member name (required)
• INLINECODE2: Role/position (required)
• INLINECODE3: Whether team Leader (required, true/false, only one Leader per team)
• INLINECODE4: Enable status true/false (required)
• INLINECODE5: Tags, comma-separated (required)
• INLINECODE6: Expertise skills, comma-separated (required)
• INLINECODE7: Weak areas, comma-separated (required)
• INLINECODE8: Whether to load workflow prompts (optional, true/false, default: true for leader, false for others)
• INLINECODE9: Group name for categorization (optional, used for organizing team members)

Reset Data

Clear all team data and reset to initial state:

CODEBLOCK3

Output:
CODEBLOCK4

⚠️ Warning: This operation is irreversible. All data in ~/.agent-team/team.json will be permanently deleted.

Data Storage

Team data is stored in ~/.agent-team/team.json, shared globally. Directory is auto-created if it doesn't exist.

Common Use Cases

Finding the right person for a task

Changing team leader

Temporarily disabling a member

Adding a new team member

Error Handling

Data file not found

Invalid JSON

Missing required parameters

Single leader constraint

References

• - Plugin Installation Guide - How to install and configure the OpenClaw plugin