From grok-swarm
Spawn an autonomous Grok agent to accomplish tasks. Use when asked to "use grok agent to refactor X", "let grok agent handle this", "grok agent mode", "autonomous grok". Triggers: "grok agent", "agent mode", "autonomous grok", "grok-swarm-agent"
How this skill is triggered — by the user, by Claude, or both
Slash command
/grok-swarm:grok-swarm-agentThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Spawn an autonomous agent powered by Grok 4.20 Multi-Agent Beta that iteratively refactors, analyzes, or modifies your codebase.
Spawn an autonomous agent powered by Grok 4.20 Multi-Agent Beta that iteratively refactors, analyzes, or modifies your codebase.
use grok agent to refactor src/auth/
grok agent mode: improve error handling in lib/
let grok swarm agent add tests to the backend
| Option | Description |
|---|---|
--apply | Actually write files (default is preview mode) |
--max-iterations N | Max agent iterations (default: 5) |
--verify-cmd CMD | Command to verify changes work |
# Preview mode - shows what would change
grok agent refactor the auth module
# Actually apply changes
grok agent refactor the auth module --apply
# With verification
grok agent add tests --apply --verify-cmd "pytest tests/"
# Analyze with agent
grok agent analyze security vulnerabilities --target ./src
/grok-swarm:setup if needed)The agent reports:
npx claudepluginhub khaentertainment/grok-multiagent-plugin --plugin grok-swarmOrchestrates Claude Code's native agents for parallel multi-domain tasks like comprehensive code analysis, feature reviews, and security audits requiring diverse expertise.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.