Awesome AI Coding Agents

A curated list of AI coding assistants, agents, and frameworks — with setup guides for connecting MCP servers to each one.

The right AI tool paired with the right data sources changes everything.

---

Contents

• IDE-Based Agents
• Terminal-Based Agents
• Autonomous Agents
• Multi-Agent Frameworks
• Code Review Agents
• Specialized Agents
• MCP Integration Guides

---

IDE-Based Agents

Agent	Editor	MCP Support	Description
Cursor	VS Code fork	Native	AI-first code editor with .cursorrules and MCP
GitHub Copilot	VS Code, JetBrains, Vim	Via extensions	Code completion and chat by GitHub
Windsurf	VS Code fork	Native	Cascade flows with deep codebase understanding
Continue	VS Code, JetBrains	Native	Open-source AI code assistant
Cody	VS Code, JetBrains	Via integrations	Context-aware assistant by Sourcegraph
Supermaven	VS Code, JetBrains	Via extensions	Ultra-fast code completion
Tabnine	Multi-editor	Via extensions	AI code completion for teams

Terminal-Based Agents

Agent	MCP Support	Description
Claude Code	Native	Anthropic's agentic coding tool — thinks, writes, tests
Aider	Via config	Terminal pair programming with git integration
OpenHands	Via tools	Autonomous software development agent
Mentat	Via config	Terminal-based AI coding assistant
GPT Engineer	Via tools	Generate full codebases from prompts
Sweep	Via GitHub	AI-powered pull request generation

Autonomous Agents

Agent	MCP Support	Description
Devin	Native	Cognition's autonomous software engineer
SWE-Agent	Via tools	Autonomous bug-fixing agent
AutoGPT	Via plugins	Self-prompting autonomous agent
AgentGPT	Via tools	Browser-based autonomous agent
MetaGPT	Via tools	Multi-agent for software development
OpenClaw	Native	Personal AI assistant agent

Multi-Agent Frameworks

Framework	Language	MCP Support	Description
CrewAI	Python	Native	Role-based multi-agent collaboration
LangGraph	Python, JS	Native	Stateful multi-agent orchestration
AutoGen	Python	Via tools	Microsoft's multi-agent framework
Semantic Kernel	C#, Python, Java	Via plugins	Microsoft's AI orchestration SDK
Google ADK	Python	Native	Google's Agent Development Kit
Dify	Multi	Native	Low-code agent builder
n8n	Multi	Via nodes	Workflow automation with AI agents

Code Review Agents

Agent	Description
CodeRabbit	AI-powered code reviews on pull requests
Codium (Qodo)	Test generation and code integrity
Sourcery	Automated Python code quality improvements
Greptile	Codebase-aware AI for code reviews

Specialized Agents

Agent	Focus	Description
v0	UI/Frontend	Generates React components from prompts
bolt.new	Full-stack	Full-stack web apps from prompts
Replit Agent	Full-stack	Build and deploy apps in-browser
Lovable	Full-stack	AI-powered web app builder
Claude Artifacts	Components	Interactive component generation
ChatGPT Canvas	Writing/Code	Collaborative editing with AI

---

MCP Integration Guides

Connecting MCP to Claude Code

# Add a memory server
echo '{"mcpServers":{"memory":{"url":"'$VINKIUS_MEM0_MCP_URL'"}}}' > .mcp.json
claude  # MCP is auto-detected
/mcp    # verify connection

Connecting MCP to Cursor

1. Open Settings → MCP
2. Add server with SSE URL from Vinkius
3. Restart Cursor

Connecting MCP to CrewAI

from crewai import Agent
from crewai_tools import MCPServerAdapter

tools = MCPServerAdapter(
    server_params={"url": "https://edge.vinkius.com/vk_live_xxx/mcp"}
).tools

agent = Agent(role="Researcher", tools=tools)

Where to Get MCP Servers

The Vinkius AI Gateway provides 2,600+ production-ready MCP servers — deploy any of them in under a minute and receive an SSE endpoint URL.

Browse: vinkius.com/en/categories

---

License

MIT — Vinkius Labs