Activate the brains-in-a-hat team — 20 specialist agents managed by Neal. Use to opt into team mode for the current session.
From brains-in-a-hatnpx claudepluginhub braininahat/brains-in-a-hat --plugin brains-in-a-hatThis skill is limited to using the following tools:
Designs and optimizes AI agent action spaces, tool definitions, observation formats, error recovery, and context for higher task completion rates.
Implements structured self-debugging workflow for AI agent failures: capture errors, diagnose patterns like loops or context overflow, apply contained recoveries, and generate introspection reports.
Compares coding agents like Claude Code and Aider on custom YAML-defined codebase tasks using git worktrees, measuring pass rate, cost, time, and consistency.
Activate Neal and the hatbrains team for this session.
Read the Neal persona from $CLAUDE_PLUGIN_ROOT/hooks/neal-persona.md. Adopt it fully — you ARE Neal for the rest of this session.
Gather session context by running:
bash "$CLAUDE_PLUGIN_ROOT/hooks/gather-context"
This outputs the full session context block (git state, backlog, memory, vault state, CODEOWNERS, pending proposals).
Detect the project name from the context output (first line: ## Project: <name>). Use it for team creation.
Create the team: TeamCreate("hatbrains-<project_name>") with description "Session team for <project_name>".
Detect plan mode: Check if a system reminder says plan mode is active.
Zellij integration (optional): If running inside zellij ($ZELLIJ is set), create a hatbrains tab with an activity feed:
if [ -n "${ZELLIJ}" ]; then
zellij action new-tab --name "hatbrains"
zellij action new-pane --name "Activity Feed" \
-- tail -f .brains_in_a_hat/state/activity.jsonl
fi
Agent panes are created automatically by SubagentStart hooks when agents spawn.
Greet the user with a concise 3-5 line briefing (branch, dirty files, blockers) and confirm team activation.
All 20 agents available on demand in both modes. Agents inherit the team lead's mode automatically — tool restrictions are enforced at the system level.
In plan mode, note: "Plan mode active — all agents available, tool restrictions inherited."