Activate a TypeScript expert subagent when building a frontend component to get specialized guidance on best practices.
Use a Kubernetes subagent to get expert help configuring and deploying containerized applications.
Engage a security subagent to audit your code for vulnerabilities and OWASP compliance issues.
Switch to a Python data engineering subagent when building data pipelines and ETL workflows.
This repository is a curated collection of over 130 specialized subagents for Claude Code, Anthropic's AI coding assistant. In Claude Code, a subagent is a pre-configured AI assistant with a specific area of expertise, essentially a set of instructions and context that tells Claude to behave as a specialist in a particular domain. Rather than asking a general-purpose AI for help, you activate a specialized subagent and get focused, expert-level assistance for that task. The collection is organized into categories covering nearly every area of software development. Core development subagents include specialists for backend APIs, frontend UI, GraphQL, microservices, and real-time systems. Language specialist subagents cover TypeScript, Python, Go, Rust, Java, Swift, PHP, C++, and many others. There are subagents for DevOps and infrastructure (Docker, Kubernetes, Terraform, CI/CD), security (penetration testing, code auditing, OWASP), AI and machine learning, data engineering, mobile development, testing, and more. You install them by copying agent configuration files either globally (available in all projects) or into a specific project folder. An interactive shell script is provided to browse categories and install or remove agents. There is also a Claude Code plugin marketplace option for one-command installation of bundled category groups. A developer using Claude Code who wants more focused, expert-level assistance for a specific language, framework, or task would use this collection to activate the right specialist agent for their current work.
Generated 2026-05-18 · Model: sonnet-4-6 · Verify against the repo before relying on details.