Brain Drin

Portable agent suite for Claude Code. 51 agents, 52 commands, 100 skills — ready to deploy on any project.

Curated from the best open-source Claude Code resources: wshobson/claude-code-agents, qdhenry/claude-commands, BMAD-METHOD, VoltAgent, and Anthropic's own documentation.

What's Inside

Agents (51)

Specialized personas that Claude adopts for focused work.

Commands (52)

Slash commands (/command) for specific workflows: onboard, feature-dev, review-pr, session-handoff, debug, refactor, security-audit, db-opt, test-gen, doc-gen, and 42 more.

Skills (100)

Reusable multi-step workflows: API scaffolding, competitor intelligence, churn analysis, chaos engineering, lead scoring, terraform repair, prompt debugging, and more. All prefixed drin-*.

Memory Compiler

Python-based session persistence (.drin/memory-compiler/). Hooks into Claude Code lifecycle to save decisions and context between sessions. Requires uv.

Note: All agents, commands, and skills live inside .claude/ (a dot-folder). GitHub hides dot-folders in its web UI — you need to clone the repo to see the full content.

Prerequisites

• Claude Code CLI installed
• Git
• uv (only for the memory compiler)

Install

Clone the repo and run the installer from your target project:

git clone https://github.com/pedrol-cmd/brain-drin.git
cd your-project
bash /path/to/brain-drin/setup.sh

The script copies agents, commands, skills, and protocol files into your project's .claude/ directory. If .claude/ already exists, it backs it up first.

For the memory compiler (optional):

cd .drin/memory-compiler && uv sync

Usage

1. Direct agent: @code-reviewer review this PR — activates the specialist
2. Slash command: /onboard — maps architecture of current project
3. Orchestrator: @team-lead plan the migration — decomposes complex tasks across agents
4. Session persistence: /session-handoff — saves context for next conversation

Read CORE-PROTOCOL.md for the full efficiency guide.

Sources

License

MIT