I wrote yesterday that Windsurf added MCP (Model Context Protocol).

The ability to use MCP is also available in Cursor starting from version 0.45. Here's a video on how to set it up:

https://www.youtube.com/watch?v=brhs5DogIf8

The video demonstrates integration with Docker and retrieving video data from YouTube.

Here are some of MCP toos that you can find at https://mcpservers.org/:

• Web scraping (Puppeteer, Brave Search, Fetch, Jina Reader)
• Communication (Slack, Bluesky)
• Productivity (Notion, Apple Shortcuts)
• Development (GitHub, GitLab, GitTools, Phabricator, Obsidian)
• Databases (PostgreSQL, Sqlite, MySQL, BigQuery, Mongo, MongoDB)
• Cloud services (Cloudflare, OpenAI, Kagi, Exa, HuggingFace Spaces)
• File system (Google Drive, Cloud Storage, Secure file operations)
• Version control (Version Control)
• Docker, Kubernetes
• Other (Sentry, Memory, Google Maps, Cloudflare, CoinCap, MetoroQuery, Windows CLI, Playwright, , Google Search Console, Pandoc, Data Exploration, any-chat-completions-mcp, Minima)

#cursor #mcp

Comments

Cursor's journey began by essentially cloning GitHub Copilot's in-editor code completion. They initially called it Copilot++ and steadily improved it, focusing on its ability to understand multiple lines of code.

By mid-2024, the feature was rebranded as Cursor Tab (a nod to its activation method by Tab button 😉) — reflecting gains in speed and accuracy. And just a month ago, they announced a new model called Fusion, which improves support for large files.

Now, competitors are scrambling to catch up.

Windsurf yesterday announced Wave 3. Beyond supporting MCP tools (Model Context Protocol, also found in Cline), they've enhanced their in-editor autocompletion model, called the feature "Tab-to-jump."

And GitHub Copilot itself isn't standing still. They currently have a feature called "Next Edit Suggestion" in preview – you need to activate it in the settings (video).

#cursor #windsurf #githubcopilot #autocomplete

Comments

https://supabase.com/

Supabase is an open-source Firebase alternative with a SQL database. They showcased an example of using their service with bolt.new on their YouTube channel:

https://www.youtube.com/watch?v=GFxOwNiioT0

Lovable also announced improved integration with Supabase:

• Less related errors
• Automatically reads edge function logs
• Customizable signup/login flows

#bolt #lovable #supabase

Comments

In MS VS Code 1.97
Copilot is already built into the program and is not installed as an extension. Voice input is still a separate extension (VS Code Speech).

The o3-mini model has been added for all plans, even the free one, which has 50 requests/month (news). However, it didn't appear automatically for me; I had to figure out the settings:

In the preview, there's a feature to set context from a Markdown file; here, you need to create it in .github/copilot-instructions.md

Also, people have discovered that you will not be able to name your variables "sexHere" and "sexThere," as Copilot intentionally stops working with code that contains predefined forbidden words (discussion).

#githubcopilot

Comments

https://youtu.be/C95drFKy4ss

The agent awakens

GitHub Copilot now looks like Cursor; they even implemented switching to agent mode in the same place. Here, Composer is called "Edit with Copilot," and you can also switch models.

Now, it can also accept a screenshot where we show what's wrong. There's also the ability to upload markdown files to provide context.

I see there's a microphone for voice input, which Cursor doesn't have...

Announcement: https://github.blog/news-insights/product-news/github-copilot-the-agent-awakens/

#githubcopilot

Comments

The website https://cursor.directory/

is a directory of .cursorrules templates
for Cursor AI and other ai-ide

covering various topics
(currently, there aren't many templates available).

In the Learn section (learn), video tutorials are compiled on diverse subjects, ranging from a basic introduction to Cursor AI to more advanced development techniques such as Composer and Rules, as well as integration with other platforms and services.

Similar websites include:

• https://dotcursorrules.com/
• https://cursorlist.com/

#cursor #promts

Comments

On the Aider LLM Leaderboards, for several days now, the first place has been occupied not by a single model, but by the combination of DeepSeek R1 and claude-3-5-sonnet-20241022.

This setup likely inspired the creation of the

DeepClaude 0.1.0
https://deepclaude.com/

Currently, DeepClaude is a very minimalist BYOK (Bring Your Own Keys) open-source system built in Rust - repo. It was created by An Asterisk as a side project and requires users to pay two providers and input their API keys. Interestingly, it’s unclear why the developers didn’t opt to integrate with OpenRouter.ai instead — perhaps someone will code this.

The system can be used on their website or deployed locally. It’s not an IDE or a plugin — just a chat interface powered by two models.

Comments

https://www.youtube.com/watch?v=gYLNxUxVomY

Main ideas from the video "I spent 400+ hours in Cursor, here's what I learned":

📂 Create a folder for instructions or prompts.
Use markdown files and comments in the code to describe the project, technology stack, technology, and database structure, rules of behavior, and important instructions. Regularly updating these files helps AI better understand the project context. Cursor does not always parse framework/library documentation well, so it's better to copy them - check the documentation yourself!

Create a file roadmap.md to track the project's status, current goals, and future steps.

Use additional tools, such as Perplexity for searching information and Whisper Flow for voice input. Speaking is much faster than typing! Separately understand your architecture and brainstorm better concepts, this can be done with a "thinking" chat like o1. Do not let AI make big decisions.

⚠️ Danger: Not understanding the code created by AI accumulates "technical debt".

(Video from January 17 regarding r1 and o3) Choosing the right AI model is important - when generating code, Sonnet 3.5 has an advantage, as it provides high accuracy. Queries need to be built in detail: describe the task, add file tags, instructions, and context.

🗣 Chat: For questions, getting information, lists.
✍️ Composer: For making changes to the code, especially complex ones, use Agent mode.

There are tricks like "the less code, the better" or "think like a senior developer", "start with three paragraphs of thoughts", "do not delete comments", to get quality results. Do not overload AI with large instructions, but break tasks into parts.

• The fewer lines of code, the better
• Proceed like a Senior Developer
• DO NOT STOP WORKING until...

add:

• v0: For quick creation of first design versions. 🎨
• Claude chat: For consultations, brainstorming. 🤔
• Lavable/Bolt: For quick creation of MVP backend (with caution, control technical debt).

mentions https://repoprompt.com/

#cursor #promts

Comments

https://github.com/block/goose

Goose1.0.4

Project Goose - an open-source artificial intelligence agent. Offers both a desktop interface and a command-line interface (CLI). It uses extensionsto connect to existing tools and applications, such as GitHub or JetBrains development environment, thanks to MCP. The catalog is still very minimalist.

Goose is multifunctional, supporting various LLM providers and capable of handling a wide range of engineering tasks, including
code migration, project onboarding with programming languages, code refactoring, performance evaluation, code coverage improvement, API framework creation, and modular test generation..

Although it was initially focused on engineering, its open nature encourages the community to explore various applications.

Fortunately, not in Python, but in Rust. It supports macOS and Linux. In Windows, only through WSL ( Windows Subsystem for Linux).

Comments

https://www.reddit.com/r/cursor/comments/1ienr72/o3mini_is_out_to_all_cursor_users/?rdt=53528
and https://codeium.com/changelog

Windsurf Cascade and Cursor added the o3-mini model from openai.

#windsurf #cursor #newllmmodel

Comments

https://www.cursor.com/changelog

Cursor

DeepSeek models: DeepSeek R1 and DeepSeek v3 are supported in versions 0.45 and 0.44. You can enable them in Settings > Models. We host these models in the USA.

But the composer (automatic code writing system) still supports only outdated GPT-4o or Claude-3.5 for now. In the chat with r1, the <think> tag is not a hidden block.

#cursor #newllmmodel

Comments

https://www.all-hands.dev/

All Hands OpenHands
(previously OpenDevin)

Open Source Python tool inspired by Devin ', launching agents for programming: writes code, executes commands, goes online.

Runs through Docker, they advise using anthropic/claude-3-5-sonnet-20241022, but you can also use Gemini and DeepSeek - configured by entering an API key. The solution is automatic with a large number of requests, so it consumes a lot of tokens.

There is a waiting list to try their hosted version.

video on how to install and configure:
https://youtu.be/WDP2S4MOXPY

Comments

https://github.com/RooVetGit/Roo-Code
plugin is actively changing from just an automatic code editor to the orchestration of various agents with their prompts and limitations.

Roo Code Update (prev. Roo Cline) v 3.3

Code Actions

Roo Code now integrates directly with the native VS Code Code Actions system, providing quick fixes and refactoring options right in the editor. Look for a light bulb above the code 💡

Advanced mode capabilities:

• Markdown Editing: implemented one of the most requested features - "Ask" and "Architect" modes can now create and edit Markdown files!

• Custom File Restrictions: in general, custom modes can now be limited to certain file patterns (can only edit Markdown files).

• Self-switching mode: modes can intelligently request switching between each other depending on the task. For example, the "Code" mode may request switching to the "Test Engineer" mode when it is ready to write tests.

#roo

Comments

https://www.technologyreview.com/2025/01/20/1110180/the-second-wave-of-ai-coding-is-here/

In the article, those who are now telling how their models and services for programmers will soon replace thousands of developers, but for now are happy to deceive at least one investor:

https://zencoder.ai/ (try for free with prices)
https://www.merly.ai/ (trial without price)
https://cosine.sh/ (waitlist)
https://www.tessl.io/ (waitlist)
https://www.poolside.ai/ (looks like B2B)

Comments

https://www.youtube.com/watch?v=itsGX3UioLk

Generating a draft site using bolt.new when the request for it is created by infranodus.com (there is a two-week trial) analyzing the gap between supply and demand, in this case, real estate in Berlin and the case of healthy food.

#bolt

Comments

https://www.youtube.com/watch?v=yHDvCGNjIqk

The video compares Bolt and Lovable, two AI tools for creating web applications.

The author gives both platforms the same task: to create a Trello-like application with draggable elements.

• Bolt generated the initial version a little faster than Lovable.
• Both applications allowed creating tasks and moving them between columns, but not editing.

Both tools quickly added this feature after the corresponding request.

When the author asked to change the design to Spotify style, both AIs successfully changed the color scheme to dark with green accents.

• Bolt in this case offered a more interesting highlighting effect when hovering over the buttons.

When adding functionality for multiple boards,

• Bolt implemented switching between them via a dropdown menu, and Lovable placed the board names in the top navigation bar, which the video author liked more.

In general, both tools are effective for rapid development, but the choice between them may depend on the user's priorities:

• Bolt may be better for those who value speed, and Lovable – for those who prefer a more intuitive "out of the box" interface

PS from me: Lovable is a closed product (formerly GPT Engineer), where 5 requests/day are free and you cannot save the code directly. Bolt.new and fork bolt.diy are available on github - you can install it on your computer or use the website bolt.new
#bolt #lovable

Comments