
Claude Army
Lets you spawn multiple Claude Code agents from chat and have them work on different codebases in parallel. You get six tools: dispatch_task to kick off an agent in a project directory (with optional team mode for automatic work decomposition), check_tasks to see what's running, get_task_output for full transcripts, get_agent_team for hierarchical views when sub-agents are involved, plus cancel_task and purge_tasks for cleanup. Built on Claude Code's stream-json output for real-time progress tracking. Useful when you want to refactor one repo while adding features to another without switching context, or when a complex task benefits from multiple specialized agents running simultaneously.
- GitHub stars
- 1
- tools
- 6
- transport
- STDIO
- last updated
- Feb 7, 2026
Tools
6 toolsPublic tool metadata for what this MCP can expose to an agent.
dispatch_taskDeploy a Claude Code agent to a project directory. Supports `mode: "team"` for multi-agent coordination
Deploy a Claude Code agent to a project directory. Supports `mode: "team"` for multi-agent coordination
No parameter schema in public metadata yet.
check_tasksMonitor status of all running/completed agents
Monitor status of all running/completed agents
No parameter schema in public metadata yet.
get_task_outputRetrieve what an agent did and its full output
Retrieve what an agent did and its full output
No parameter schema in public metadata yet.
get_agent_teamDetailed view of a lead agent and all its sub-agents
Detailed view of a lead agent and all its sub-agents
No parameter schema in public metadata yet.
cancel_taskStop a running agent gracefully
Stop a running agent gracefully
No parameter schema in public metadata yet.
purge_tasksClean up completed/failed tasks
Clean up completed/failed tasks
No parameter schema in public metadata yet.
Sponsors
Readme
πͺ ClaudeArmy
An MCP server that lets Claude orchestrate multiple Claude Code agents working autonomously across different projects.
ClaudeArmy Demo
What It Does
ClaudeArmy gives Claude (in the chat interface) the ability to spawn background Claude Code processes that work independently on coding tasks. Think of it as a chain of command:
- You β give strategic direction
- Claude (chat) β breaks it down and dispatches agents
- Claude Code agents β execute autonomously, spawning their own sub-agents as needed
This means you can kick off work across multiple projects simultaneously, keep chatting normally, and check in on progress whenever you want.
Tools
Details
- Repository
- github.com/robbhimself-1337/claude-army
- Category
- Communication & Messaging
- Package
claude-army- Transport
- STDIO
- Registry
- MCP registry, active