This repository provides a collection of specialized AI subagents and slash commands designed to enhance development workflows within Claude Code. It offers domain-specific expertise and automation capabilities for developers, aiming to streamline tasks from code review and debugging to infrastructure management and data analysis.

How It Works

The project organizes AI capabilities into distinct "subagents" and "commands," both defined in Markdown files with frontmatter metadata. Subagents are designed for automatic invocation based on context or explicit user requests, acting as specialized AI personas. Commands are executable scripts triggered by slash commands, automating specific tasks. This modular approach allows for easy expansion and customization of Claude Code's functionality.

Quick Start & Requirements

• CLI Tool: Install globally via npm install -g bwc-cli. Use bwc init, bwc add --agent <name>, bwc add --command <name>, bwc list, and bwc search for management.
• Manual Installation: Clone the repository and copy .md files from subagents/ to ~/.claude/agents/ and from commands/ to ~/.claude/commands/.
• Prerequisites: Claude Code installed and configured. Access to ~/.claude/agents/ and ~/.claude/commands/ directories.
• Web UI: Available at buildwithclaude.com for browsing, searching, and installing.

Highlighted Details

• Over 43 specialized subagents covering development, language expertise, infrastructure, security, data science, and more.
• More than 39 community-contributed slash commands for version control, code analysis, documentation, and project management.
• Includes a dedicated CLI tool (bwc-cli) for simplified installation and management of subagents and commands.
• Features a Web UI for interactive browsing, searching, and previewing of all available agents and commands.

Maintenance & Community

The project is actively maintained by Dave Poon. Contributions for new subagents and commands are welcomed via pull requests.

Licensing & Compatibility

This collection is provided under the MIT License.

Limitations & Caveats

Subagent invocation relies on Claude Code's internal context analysis; explicit invocation may be necessary for precise control. Command usage requires adherence to specific syntax and argument formats.