- The AI Stack for devs
- Posts
- From figma to production
From figma to production
5 MCP servers that do the boring work for you
Sanket Gawas
September 19, 2025 • Estimated Reading Time: 8 minutes
In partnership with
Before you jump in, here’s something you’ll love.
I’ve put together a curated list of MCP servers and MCP registries to support your development workflows. When you subscribe to the newsletter, you’ll get instant access to this resource in your welcome email.
Bonus: I’m also working on a collection of free educational resources for building AI agents at every skill level. Coming soon for subscribers. Stay tuned!
Last week, I talked about Claude Code users experiencing major frustrations on Reddit and X, leading many to move over to Codex CLI.
Today, Claude has published an official postmortem explaining the causes behind their model’s decline.
Let’s dive into the meaty stuff now. There has been a surge in even more helpful MCP servers since I listed my top 5 must-have MCP servers for developers last month. We dive some of the good ones today.
1. Markitdown MCP
Markitdown (by Microsoft) is a Python utility and MCP server that converts dozens of file formats, PDF, Word, Excel, images, HTML, audio, even YouTube URLs directly into markdown. It’s designed for LLM pipelines, preserving document structure but optimizing for token efficiency.
Why I Like It
- Converts nearly any doc to markdown with CLI, Python, or Docker support
- Can be embedded as an MCP server for agents.
- Streams content via HTTP, STDIO, or Docker for scalable usage
- Extensible with plugin architecture
- Useful for indexing docs into retrieval systems.
Use Case
Pipe onboarding guides, engineering wikis, or design docs directly into your agent’s context.
2. Serena MCP
Serena gives your coding agent a real IDE brain. Instead of just regex search or token guessing, it hooks into language servers so the agent can navigate and refactor code with symbolic understanding.
Consider it as providing Claude Code or Cursor with the ability to see your codebase like an experienced developer does: by finding references, searching across symbols, and editing in a context-aware manner.
Why I Like It
- Exposes tools for symbol-level code search, reference lookups, and semantic editing.
- Boosts token and cost efficiency.
- Integrates with a huge range of languages: Python, TS/JS, Go, Rust, Java, C/C++, Swift, Kotlin and more.
- Custom modes/contexts for IDEs, agent frameworks, or planning/analysis use-cases.
- Ideal for navigating large monorepos or multi-language codebases.
3. Sentry MCP
Sentry MCP connects your agent directly to Sentry, letting it retrieve, analyze, and triage application errors and performance issues at the code or org level.
Why I Like It:
- Remote server acts as middleware to Sentry’s API
- Supports both production cloud and self-hosted Sentry (via STDIO or HTTP)
- AI-powered search tools (find events/issues with natural language) with OpenAI key
- Auth/OAuth via Sentry’s native user tokens and App keys
- Focused on developer and debugging workflows — not just generic Sentry API access
Use Case:
Have your agent summarise the latest critical errors, generate auto-triage plans, or even recommend fixes using Sentry context.
4. Postman MCP
Postman MCP lets AI agents, assistants, and chatbots manage Postman collections, environments, and workspaces automating API workflows programmatically.
Why I Like It:
- Plug-and-play authentication with your Postman API key
- Enables rich automation: create collections, run requests, manage environments, edit variables, orchestrate APIs
Use Case:
Delegate repetitive API tasks like updating environments, running regression requests, or spinning up collections to the agent.
5. Figma Dev Mode MCP
Figma Dev Mode MCP bridges the gap between design and development, letting agents extract variables, generate code, and map components directly from Figma files.
Why I Like It:
- Generates code (default: React + Tailwind) from Figma frames/selections via MCP
- Extracts design tokens, layouts, screenshots, and design system rules
- Integrates with VS Code, Cursor, Claude, and more via Streamable HTTP
Use Case:
Automate the conversion of Figma designs into code, and enforce your design system out of the box.
The Simplest Way To Create and Launch AI Agents
Imagine if ChatGPT and Zapier had a baby. That's Lindy.
With Lindy, you can build AI agents in minutes to automate workflows, save time, and grow your business. From inbound lead qualification to outbound sales outreach and web scraping agents, Lindy has hundreds of AI agents that are ready to work for you 24/7/365.
Stop doing repetitive tasks manually. Let Lindy's agents handle customer support, data entry, lead enrichment, appointment scheduling, and more while you focus on what matters most - growing your business.
Join thousands of businesses already saving hours every week with intelligent automation that actually works.
If you’re not a subscriber, here’s what you missed this month
Subscribe to get access to such posts every week in your email.
🚢 What’s shipping this week?
Vercel AI SDK tools now provides state management, debugging tools, and structured artefact streaming for building advanced AI interfaces.
Github released their MCP server Registry this week
👓️ Worth reading for Devs
🤠 What you may have missed last month
My Recommendations
Techpresso gives you a daily rundown of what's happening in tech and read by 300,000+ professionals.
The Deep View The go to daily newsletter for 250k+ founders and knowledge workers who want to stay up to date with artificial intelligence.
Looking for more such updates in your inbox? Discover other newsletters that our audience loves to read here