claude code sub agents

Report Abuse

Basic Information

This repository is a curated collection of specialized subagents designed to extend Claude Code for software development workflows. It bundles 33 domain-specific agent definitions organized into categories such as development, infrastructure, quality & testing, data & AI, security, specialization, and business. Each subagent is a Markdown document that defines a role, trigger conditions, expertise areas, capabilities, and a system prompt so Claude Code can automatically invoke or explicitly call the right specialist. The collection includes a master orchestrator called agent-organizer for assembling multi-agent teams and coordinating multi-step tasks. The README documents installation by placing agent files under the user's Claude agents directory, optional MCP server configurations for enhanced capabilities, standardized subagent file format, usage patterns (automatic, explicit, hybrid), example orchestrations, and contribution guidelines for adding new agents.

Links

Categorization

App Details

Features
The repository emphasizes intelligent auto-delegation where Claude Code selects optimal subagents by context analysis and task classification. It provides domain expertise agents for frontend, backend, languages, devops, performance, QA, data, ML, security, and documentation. Multi-agent orchestration patterns are documented including sequential, parallel, validation, and iterative flows. A meta-orchestrator (agent-organizer) can analyze projects, assemble focused teams, and manage quality gates. The README lists MCP server integrations to enable sequential thinking, up-to-date context access, UI generation, browser automation, and filesystem tools. Each subagent follows a standardized frontmatter and system prompt format for consistent behavior. The repo includes usage examples, resource metrics for example runs, troubleshooting guidance, and contribution/quality standards to ensure integration-ready agent definitions.
Use Cases
For developers and teams using Claude Code, this collection accelerates common engineering tasks by providing ready-made specialist agents to design, implement, test, review, secure, and optimize code and infrastructure. It reduces overhead by enabling automatic or explicit delegation to experts like backend-architect, frontend-developer, code-reviewer, test-automator, performance-engineer, and security-auditor. The agent-organizer supports complex multi-phase projects by assembling small focused teams and orchestrating handoffs and validations. Installation is simple: copy agent Markdown files into the Claude agents directory to make them available. The README includes practical examples of end-to-end feature implementation and large system design to illustrate expected token usage, execution flow, and deliverables. Contribution guidance and standardized formats make it straightforward to add or customize subagents for project-specific workflows.

Please fill the required fields*