From claw
Add a new agent to a running Claw workspace. Use when the user wants another AI agent in their team without disrupting existing ones.
How this skill is triggered — by the user, by Claude, or both
Slash command
/claw:add-agentThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Add a new agent to an existing Claw workspace without disrupting running agents.
Add a new agent to an existing Claw workspace without disrupting running agents.
Find the claw workspace and read current claw.yaml to see existing agents.
Ask the user (or parse from their prompt):
${CLAUDE_PLUGIN_ROOT}/examples/ or write custom)Write an identity file for the new agent if the user wants a custom persona.
Add the new agent to the agents: list in claw.yaml.
Regenerate the stack:
cd <workspace> && ./init.sh
This preserves existing agent data while adding the new agent's directories.
cd <workspace>/generated && docker compose up -d claw-provision agent-<name>
The provision container skips existing accounts and only registers the new one.
cd <workspace>/generated && docker compose ps agent-<name>
Note: Existing agents won't know about the new bot for smart respond_to mode until restarted. If needed: docker compose restart agent-<existing-name> to update their BOT_USERS list.
npx claudepluginhub jeff-mccoy/silver-octo-meme --plugin clawGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.