Awesome AI Agents › Integration and Extensions
This repository is a curated set of subagent definitions for Claude Code, each written as a markdown file that gives the assistant domain-specific instructions for one area of software work. Claude Code can select an agent automatically from the context of a request or the user can invoke one explicitly. The agents are grouped into directories by concern. The development group covers frontend and interface work with a frontend developer, user interface and user experience designers, and React and Next.js specialists; backend work with an API and microservice architect and a full-stack developer; language specialists for Python, Go and TypeScript; platform work for mobile and Electron applications; and developer experience roles for tooling improvement and legacy modernization. The infrastructure group contains cloud architecture, deployment and CI/CD configuration, incident response for production and deployment failures, and performance profiling and caching. The quality and testing group provides code review, architectural consistency review, quality assurance strategy, test suite creation across unit, integration and end-to-end levels, and debugging. The data and AI group covers ETL pipelines and warehouses, data analysis and SQL, database and PostgreSQL optimization, GraphQL schema design, LLM application and RAG construction, machine learning pipelines and model serving, and prompt optimization. Security is represented by an auditor focused on vulnerabilities and OWASP compliance, specialization by an OpenAPI documenter and a technical writing expert, and business by a product manager. Above these sits an agent organizer that acts as a meta-orchestrator: it analyzes project requirements, detects the technology stack and architecture patterns, assembles a small team of one to three agents suited to the task, and manages the collaborative workflow between them. The collection is intended for developers who want ready-made role definitions instead of writing their own.
https://github.com/lst97/claude-code-sub-agents
Composio is a production-ready toolset that equips AI agents and LLMs with over 250 high-quality integrations across software, OS operations, and search capabilities, featuring managed authentication and a pluggable architecture for custom extensions.
A community preset pack for DeepSeek Harness that starts a session under a minimal tool schema with no injected context, then promotes to the full standard toolset once the session is durable.
Cloudflare official collection of Agent Skills, slash commands and MCP server references for building Workers, Durable Objects, Sandboxes and Agents SDK applications with coding agents.
Repository of Agent Skills for Supabase development, packaged as folders of instructions and reference files that AI coding agents discover and load when working with Supabase or Postgres.
Configuration repository for Claude Code distributing a plugin marketplace of skills and agents covering deep research, image generation, GitHub automation and translation, plus multi-model gateway settings.
A Claude Code skill of 27 scripts that build and test Xcode projects and drive the iOS Simulator through accessibility-based navigation with token-efficient, progressively disclosed output.
A curated list of community plugins for DeepSeek Harness, the plugin-based open-source coding agent, sorted into categories such as models, tools, memory, skills, workflow, security and remote access.
A modular collection of 20+ tools, function pipes and filters that extend Open WebUI with search, media generation, semantic routing and autonomous planner agents.