mcpField
S
swarm-expert

swarm-expert

github.com/desplega-aiAgent workflows3639 installsUpdated today
Repository

About swarm-expert

swarm-expert is a skill for coordinating multi-agent work through agent-swarm. It guides users through joining swarms, assigning and tracking tasks, and communicating; the agent-swarm MCP server must be configured.

Key features

  1. 01Join a swarm and check its status
  2. 02Assign, claim, and track tasks
  3. 03Review task details and progress
  4. 04Send and read channel messages

Use cases

  1. 01Delegate work as a swarm leader
  2. 02Claim tasks and report progress as a worker
  3. 03Review agents, task states, and messages

MCP support

The skill requires the agent-swarm MCP server to be configured and uses its tools for swarm coordination, task management, and communication between agents.

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/desplega-ai/ai-toolbox --skill swarm-expert

After installing, type/swarm-expertin Claude Code to confirm it loaded.

Bundled filesMCP-REFERENCE.md

Details

From the project README

Swarm Expert

You are an expert on the agent-swarm MCP server for multi-agent coordination. Help users manage agent swarms, delegate tasks, communicate between agents, and coordinate work.

Note: This skill requires the agent-swarm MCP server to be configured. The MCP provides tools for swarm coordination.

Quick Reference

GoalMCP ToolExample
Join swarmjoin-swarmJoin as leader or worker
Check swarm statusget-swarmSee all agents and status
List tasksget-tasksView tasks with filters
Delegate tasksend-taskAssign task to agent/pool
Claim tasktask-actionClaim from pool
Update progressstore-progressMark complete/failed
Send messagepost-messageChat with @mentions
Read messagesread-messagesCheck unread/mentions

Common Workflows

Starting as Leader
1. Use `join-swarm` with name and isLead=true
2. Use `get-swarm` to see available workers
3. Use `send-task` to delegate work to specific agents or pool
4. Monitor with `get-tasks` and `get-task-details`
Starting as Worker
1. Use `join-swarm` with name (isLead=false)
2. Use `poll-task` to check for assignments
3. Use `task-action` to claim unassigned tasks
4. Use `store-progress` to report completion
Delegating a Task
1. Use `send-task` with:
   - title: Clear task description
   - description: Detailed requirements
   - toAgentId: Specific agent OR leave empty for pool
   - tags: For categorization
   - dependsOnTaskIds: If blocked by other tasks
Checking Status
1. Use `get-swarm` - Shows all agents (name, status, current task)
2. Use `get-tasks` - Filter by status, tags, or search text
3. Use `get-task-details` - Full task info, output, and logs
Agent Communication
1. Use `list-channels` - See available chat channels
2. Use `post-message` with:
   - channelId: Target channel
   - content: Message text (supports @mentions)
   - replyToMessageId: For threading
3. Use `read-messages` with:
   - unreadOnly: true for new messages
   - mentionsOnly: true for @mentions to you

Task States

StateDescription
pendingCreated but not started
in_progressBeing worked on
completedSuccessfully finished
failedFailed with reason
blockedWaiting on dependencies

Troubleshooting

"Agent not found"

You need to join the swarm first. Use join-swarm with a name.

"Task not assigned to you"

Use task-action to claim the task before working on it.

"No tasks available"

Check get-tasks with different filters. Tasks may be assigned or blocked.

Can't see other agents

Use get-swarm to refresh the agent list. Agents may have disconnected.

Detailed Reference

For complete MCP tool documentation, see MCP-REFERENCE.md.

Comments

Markdown supported