claude-sub-agents

What problem does it solve?

General-purpose AI can struggle with highly specialized tasks or maintaining focus in complex conversations. Subagents provide task-specific expertise and isolated contexts, improving efficiency and accuracy.

Core Features & Use Cases

• Specialized Expertise: Configure subagents with custom system prompts and tools for specific domains (e.g., code reviewer, debugger, data scientist).
• Context Preservation: Each subagent operates in its own context window, preventing main conversation pollution.
• Flexible Configuration: Define subagents via Markdown files or CLI, with options for tool access and model selection.
• Use Case: When you need a thorough code review, explicitly invoke the code-reviewer subagent. It will operate in its own context, focusing solely on code quality and security, and provide structured feedback without distracting the main conversation.

Quick Start

Use the claude-sub-agents skill to view all available subagents and create a new project-level subagent for test automation.