team-communication-protocols
π―SkillRules for how AI agents on the same team should talk to each other. It covers when to message one teammate or everyone, how a lead approves a teammate's plan, and how to shut the team down cleanly when work is done.
Installation
npx skills add wshobson/agents --skill team-communication-protocolsRun it in a terminal; it asks which AI coding tools to install it for (Claude Code, Codex, Cursor and more).
Uses skills.sh Β· Node.js 22.20+
After installing
Ask for related work and the agent uses it on its own, or call it yourself.
/team-communication-protocolsInstall by hand
Copy the whole skill folder from GitHub to the place below. Copying only SKILL.md can leave out files the skill uses.
Open the skill folder β- This project
.claude/skills/team-communication-protocols/ - All projects
~/.claude/skills/team-communication-protocols/
These paths are for Claude Code. For other tools, the command above puts it in the right place.
What this skill does
Overview
This skill sets clear rules for how agents on an AI agent team send messages to each other. It explains which kind of message to use and when. It also covers plan approval and shutting the team down at the end.
Key Features
- Send a direct message to one teammate for normal updates and questions.
- Save broadcasts, messages to every teammate, for urgent changes that affect everyone.
- Have a lead review and approve or reject a teammate's plan before work starts.
- Shut teammates down safely, and wait if one is still working.
- Fix common problems, such as a silent teammate or two teammates waiting on each other.
Who is this for?
This is for people who run teams of AI agents that work together on code. It helps when agents talk too much, miss key handoffs, or get stuck. Following these rules keeps the team calm, cheap to run, and on track.
Summarized by AI from this skill's SKILL.md. Suggest a better one
SKILL.md
View on GitHub βStructured messaging protocols for agent team communication including message type selection, plan approval, shutdown procedures, and anti-patterns to avoid. Use this skill when establishing communication norms for a newly spawned team, when deciding whether to send a direct message or a broadcast, when a team-lead needs to review and approve an implementer's plan before work begins, when orchestrating a graceful team shutdown after all tasks are complete, or when debugging why teammates are not coordinating correctly at integration points.
Team Communication Protocols
Protocols for effective communication between agent teammates, including message type selection, plan approval workflows, shutdown procedures, and common anti-patterns to avoid.
When to Use This Skill
- Establishing communication norms for a new team
- Choosing between message types (message, broadcast, shutdown_request)
- Handling plan approval workflows
- Managing graceful team shutdown
- Discovering teammate identities and capabilities
Message Type Selection
message (Direct Message) β Default Choice
Send to a single specific teammate:
{
"type": "message",
"recipient": "implementer-1",
"content": "Your API endpoint is ready. You can now build the frontend form.",
"summary": "API endpoint ready for frontend"
}
Use for: Task updates, coordination, questions, integration notifications.
broadcast β Use Sparingly
Send to ALL teammates simultaneously:
{
"type": "broadcast",
"content": "Critical: shared types file has been updated. Pull latest before continuing.",
"summary": "Shared types updated"
}
Use ONLY for: Critical blockers affecting everyone, major changes to shared resources.
Why sparingly?: Each broadcast sends N separate messages (one per teammate), consuming API resources proportional to team size.
shutdown_request β Graceful Termination
Request a teammate to shut down:
{
"type": "shutdown_request",
"recipient": "reviewer-1",
"content": "Review complete, shutting down team."
}
The teammate responds with shutdown_response (approve or reject with reason).
Communication Anti-Patterns
| Anti-Pattern | Problem | Better Approach |
|---|---|---|
| Broadcasting routine updates | Wastes resources, noise | Direct message to affected teammate |
| Sending JSON status messages | Not designed for structured data | Use TaskUpdate to update task status |
| Not communicating at integration points | Teammates build against stale interfaces | Message when your interface is ready |
| Micromanaging via messages | Overwhelms teammates, slows work | Check in at milestones, not every step |
| Using UUIDs instead of names | Hard to read, error-prone | Always use teammate names |
| Ignoring idle teammates | Wasted capacity | Assign new work or shut down |
Plan Approval Workflow
When a teammate is spawned with plan_mode_required:
- Teammate creates a plan using read-only exploration tools
- Teammate calls
ExitPlanModewhich sends aplan_approval_requestto the lead - Lead reviews the plan
- Lead responds with
plan_approval_response:
Approve:
{
"type": "plan_approval_response",
"request_id": "abc-123",
"recipient": "implementer-1",
"approve": true
}
Reject with feedback:
{
"type": "plan_approval_response",
"request_id": "abc-123",
"recipient": "implementer-1",
"approve": false,
"content": "Please add error handling for the API calls"
}
Shutdown Protocol
Graceful Shutdown Sequence
- Lead sends shutdown_request to each teammate
- Teammate receives request as a JSON message with
type: "shutdown_request" - Teammate responds with
shutdown_response:approve: trueβ Teammate saves state and exitsapprove: false+ reason β Teammate continues working
- Lead handles rejections β Wait for teammate to finish, then retry
- After all teammates shut down β Call
TeamDeleteto remove team resources
Handling Rejections
If a teammate rejects shutdown:
- Check their reason (usually "still working on task")
- Wait for their current task to complete
- Retry shutdown request
- If urgent, user can force shutdown
Teammate Discovery
Find team members by reading the config file:
Location: ~/.claude/teams/{team-name}/config.json
Structure:
{
"members": [
{
"name": "security-reviewer",
"agentId": "uuid-here",
"agentType": "team-reviewer"
},
{
"name": "perf-reviewer",
"agentId": "uuid-here",
"agentType": "team-reviewer"
}
]
}
Always use name for messaging and task assignment. Never use agentId, role names, or unsuffixed aliases directly. If a teammate was spawned as team-lead-2, send to team-lead-2, not team-lead.
Troubleshooting
A teammate is not responding to messages. Check the teammate's task status. If it is idle, it may have completed its task and is waiting to be assigned new work or shut down. If it is still active, it may be mid-execution and will process messages once the current operation finishes.
A teammate says it cannot see SendMessage.
Check the teammate agent's tools: frontmatter. Agent Teams communication tools such as SendMessage, TaskList, TaskGet, and TaskUpdate must be listed explicitly when an agent uses a restricted tool allowlist.
The lead is sending broadcasts for every status update.
This is a common anti-pattern. Broadcasts are expensive β each one sends N messages. Use direct messages (type: "message") for point-to-point updates. Reserve broadcasts for critical shared-resource changes like an updated interface contract.
A teammate rejected a shutdown request unexpectedly.
The teammate is still working. Check the rejection reason in the shutdown_response content field, wait for the work to finish, then retry. Never force-terminate a teammate that has unsaved work.
A plan_approval_request arrived but the request_id is missing.
The teammate called ExitPlanMode without the required request context. Have the teammate re-enter plan mode, complete exploration, and call ExitPlanMode again. The request_id is generated automatically by the plan mode system.
Two teammates are waiting on each other and neither is making progress. This is a deadlock: both are blocked waiting for the other to finish first. The lead should send a direct message to one teammate with a stub or partial result so it can unblock and proceed.
Related Skills
- team-composition-patterns β Select agent types and team size before establishing communication norms
- parallel-feature-development β Use communication protocols to coordinate integration handoffs between parallel implementers
More from this repository10
Helps you build a consistent look for your app with Tailwind CSS v4. It sets up shared colors, sizes, dark mode and reusable components so every screen matches.
A guide to TypeScript's advanced type features, which describe the shape of your data so mistakes are caught before the code runs. It covers reusable types, types that change based on rules, and built-in helper types, with short examples.
Guidance for building reliable server-side apps with Node.js, using Express or Fastify. It covers login, error handling, databases, API design and a list of proven best practices.
Intelligent automation and multi-agent orchestration for Claude Code
Claude Code plugin for AI-powered framework migration workflows, coordinating specialized agents to plan and execute transitions between technology stacks as part of the wshobson/agents marketplace.
A Claude Code plugin for payment system integration covering Stripe, PayPal, and billing platforms with checkout flows, subscription management, and webhook handling in the wshobson/agents ecosystem.
A Claude Code plugin bundling core development tools and agents for everyday tasks like debugging and code generation, part of the wshobson/agents marketplace with 73 focused plugins.
Claude Code plugin for observability and monitoring including logging, metrics, distributed tracing, and alerting, part of the wshobson/agents marketplace with 73 focused plugins.
A Claude Code plugin providing specialized AI agents for deployment strategy planning and execution, covering blue-green deployments, canary releases, rolling updates, and zero-downtime patterns.
Claude Code plugin for JVM language development including Java, Kotlin, and Scala with specialized agents for idiomatic patterns, build tools, and runtime optimization in the wshobson/agents marketplace.