AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers Daniel Sogl @sogldaniel Software architect 

Transform Your Workflow with Coding Tools and MCP Servers AI-Powered Development 

Daniel Sogl • Software architect @ Thinktecture AG • Focus: Angular, Capacitor and AI-Coding • Socials: https://linktr.ee/daniel_sogl About me Transform Your Workflow with Coding Tools and MCP Servers AI-Powered Development 

• AI is transforming software development workflows • Tools like GitHub Copilot and Claude promise increased productivity • Challenges: too many tools, unclear benefits, rapid change AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers Why AI Matters in Development 

• Developers' complete tasks 26% faster (MIT, 2023) • Coding tasks on average 55% faster (GitHub, 2024) • Research work reduced by 67% - “Dead Internet Theory” (JetBrains, 2024) • Less burnout, more time for creative tasks (GitHub, 2024) AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers Productivity claims over productivity claims 

Transform Your Workflow with Coding Tools and MCP Servers AI-Powered Development 

Stay focused • Code Completion Assistants (e.g., Copilot, Gemini Code) • AI IDEs (e.g., Cursor, Windsurf) • CLIs (e.g., Claude Code, CodeX, Gemini CLI) • Chat Assistants (e.g., Claude Desktop, Copilot Chat) • Agentic Tools and MCP Servers • Full-Stack Builders (e.g., Lovable.dev) AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers Tool Categories 

Full-stack applications with a user-friendly approach • User friendly interface for nontechnical users • Uses React • Chat mode to plan before acting • Strong Supabase integration for authentication and databases • 3rd Party integrations like Stripe, Resend or OpenAI • Optimized for team collaboration with GitHub- first workflows Tool Spotlight: Lovable.dev Transform Your Workflow with Coding Tools and MCP Servers AI-Powered Development 

The all-inclusive AI Coding tool • Can be used across multiple IDEs (VS-Code, IntelliJ, Xcode etc.) • Free to use / 10$ per month for the pro version • Provides ask, edit and agent mode • Inline autocompletion, code reviews and test generation • Provides different LLMs (GPT, Gemini, Sonnet) • MCP (Model Context Protocol) support Tool Spotlight: GitHub Copilot – Edit/Agent Mode Transform Your Workflow with Coding Tools and MCP Servers AI-Powered Development 

• Access Google Gemini 2.5 through your terminal • Open-Source • 1.000 requests per day for free • Customizable instructions and prompts • MCP support • Can be integrated into existing workflows like CI/CD environments AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers Gemini CLI 

• Chat interface with Claude AI on your desktop • Powerful for architecture, planning, and refactoring • Can write code, perform research, and search the web • Responses styles can be configured (informative, professional, funny) • Extended via MCP for local file access and actions AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers Claude Desktop 

The key for ”intelligent” workflows • Open-source protocol invented by Anthropic • Provides a consistent way for LLMs to interact with external resources • Works like the USB-C standard • Official servers are available for GitHub, Atlassian, Playwright, Stripe, Databases and more • It’s the key to useful AI-coding setups in complex environments Source: https://www.dailydoseofds.com/p/visual-guide-to-model-context-protocol-mcp/ AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers Model Context Protocol (MCP) 

Understand the limits of LLMs' Mistakes • LLMs are not intelligent by default • LLMs were trained on older data • LLMs have different context windows • Not all LLMs are optimized for coding Solutions • Keep track of LLM benchmarks for coding (SWE-Bench) • Compare pricing for optimized token per dollar usage • Give LLMs access to unknown context • Add custom instructions AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers Common mistakes when working with AI-Coding tools 

Don’t try to build entire features with a single prompt Mistakes • It's fascinating to see how fast AI-tools can generate code and good results • Developers often try to solve complex features with overly simple prompts • Agent modes can try to find needed context to solve the task Solutions • Split tasks into smaller tasks • Use MCP-Server like sequential-thinking for splitting • Write user stories that can be broken down AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers Common mistakes when working with AI-Coding tools 

Write precise prompts Mistakes • LLMs' are not intelligent by default • LLMs' where trained on older data • LLMs' have different context windows • Not all LLMs are optimized for coding Solutions • Keep track of LLM benchmarks for coding (SWE-Bench) • Compare pricing for optimized token per dollar usage • Give LLMs' access to unknown context AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers Common mistakes when working with AI-Coding tools 

The wrong MCP setup Mistakes • AI-Tools like Copilot have MCP-Tool limits • MCP-Servers can be insecure by not following authentication security standards • MCP-Servers can’t store memories by default Solutions • Use official MCP-Servers (GitHub, Atlassian, Notion etc.) • Disable unused tools • Tag tools manually (works great in Copilot) AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers Common mistakes when working with AI-Coding tools 

• sequential-thinking → splits tasks into smaller ones • GitHub MCP / Atlassian MCP → access your task dashboard, create/update tasks and review code • Database MCP-Servers → allow LLMs' to understand your database schemas • Playwright-MCP → let LLMs' interact with web applications • Perplexity-MCP → scrap web sites • Context7-MCP → fetch full documentations AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers MCP-Servers I can recommend 

• AI assisted coding tools can overwhelm developers • Generated code often is not reviewed well • Developers tend to just copy generated results without understanding the used code • LLMs' need your help to access required context to solve your tasks • Break down tasks and work iteratively • Create custom instructions and reusable prompts • Add external context or advanced thinking strategies using MCP-Servers AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers Conclusion 

• GitHub Copilot Setup: https://github.com/danielsogl/copilot-workflow-demo • Microsoft MCP-Server: https://github.com/microsoft/mcp • Context7: https://context7.com • Sequential Thinking MCP: https://github.com/modelcontextprotocol/servers/tree/main/src/sequentialthinking • Atlassian Remote MCP: https://www.atlassian.com/platform/remote-mcp-server • Gemini CLI: https://github.com/google-gemini/gemini-cli AI-Powered Development Transform Your Workflow with Coding Tools and MCP Servers Sources & Demos 

Thanks for your attention Daniel Sogl @sogldaniel [email protected]