home / skills / yeachan-heo / oh-my-claudecode / swarm

swarm skill

safe

This skill routes swarm invocations to the Team skill, enabling coordinated multi-agent task execution with familiar interfaces.

npx playbooks add skill yeachan-heo/oh-my-claudecode --skill swarm

Review the files below or copy the command above to add this skill to your agents.

Files (1)

SKILL.md

665 B

---
name: swarm
description: N coordinated agents on shared task list (compatibility facade over team)
---

# Swarm (Compatibility Facade)

Swarm is a compatibility alias for the `/oh-my-claudecode:team` skill. All swarm invocations are routed to the Team skill's staged pipeline.

## Usage

```
/oh-my-claudecode:swarm N:agent-type "task description"
/oh-my-claudecode:swarm "task description"
```

## Behavior

This skill is identical to `/oh-my-claudecode:team`. Invoke the Team skill with the same arguments:

```
/oh-my-claudecode:team <arguments>
```

Follow the Team skill's full documentation for staged pipeline, agent routing, and coordination semantics.

Overview

This skill provides a compatibility alias named "swarm" that routes invocations to the Team skill's staged pipeline. It preserves the same semantics and behavior as the Team skill while offering a familiar name for workflows built around the concept of a coordinated swarm of agents. Use it when you want the same multi-agent orchestration under the "swarm" command surface.

How this skill works

Swarm forwards any invocation to the Team skill with identical arguments, so all orchestration, agent routing, and staged-pipeline logic are handled by the Team implementation. You can pass an explicit agent count and type or just a task description; the Team pipeline receives and executes the request. Because swarm is a compatibility facade, it does not duplicate logic but acts as a stable alias to the Team orchestration layer.

When to use it

When you already use the "swarm" command in scripts or docs and want compatibility with the Team orchestration.
When you need coordinated, parallel work across N agents using the Team staged pipeline.
When you want to specify agent count and role type succinctly for a shared task list.
When integrating with Claude Code workflows that expect a swarm-style command surface.
When you need a simple alias to avoid changing existing automation that targets "swarm".

Best practices

Pass the agent count and type explicitly for predictable routing (e.g., N:agent-type "task").
Keep task descriptions concise and action-oriented to improve agent coordination and outcomes.
Rely on the Team skill documentation for details about staging, handoffs, and conflict resolution.
Use swarm for compatibility only; consult Team behavior when debugging orchestration issues.
Instrument calls in automation to capture which staged steps and agents executed each task.

Example use cases

Run 5 review agents on a codebase: /oh-my-claudecode:swarm 5:reviewer "audit auth module for security issues"
Kick off a mixed-role swarm for triage and fixes: /oh-my-claudecode:swarm 3:triage "classify and assign incoming bug reports"
Maintain backward-compatible CI scripts that currently invoke the swarm alias.
Quickly parallelize a documentation task across multiple writer agents using the alias.
Experiment with multi-agent strategies without changing existing team-oriented automation.

FAQ

Is swarm different from Team?

No. Swarm is a compatibility alias that forwards requests to the Team skill and uses the same staged pipeline and coordination semantics.

What command formats are supported?

You can call swarm with an explicit agent count and type (N:agent-type "task description") or with just a task description; both are routed to Team.