Swarm

Dispatches many independent items in parallel: create a table, fan out to subagents, aggregate results. One row = one unit of work.

概要

Swarm is a SKILL.md-based agent skill sourced from langchain-ai/langchain-skills. It is categorized under security and is listed for Codex, Claude. The source description focuses on: Dispatches many independent items in parallel: create a table, fan out to subagents, aggregate results. One row = one unit of work. AIToolly summarizes this page as a directory entry rather than copying the full third-party skill content, so users can evaluate the source, compatibility, and practical fit before installing it.

ユースケース

Evaluate Swarm before adding it to an AI agent workflow.
Use Swarm as a starting point for repeatable security tasks.
Compare Swarm with related skills from langchain-skills and other GitHub repositories.

導入方法

# Review source first
open https://github.com/langchain-ai/langchain-skills/blob/main/config/skills/swarm/SKILL.md

Copy or clone the skill folder into your agent skills directory after reviewing its instructions and scripts.

セキュリティ

Review the source SKILL.md, referenced scripts, permissions, and external services before installing Swarm. Treat third-party skills like code dependencies, especially when they can read files, call APIs, or run commands.

関連Skills