Compared to other AI models from companies like OpenAI and Google, I keep returning to Anthropic's Claude model for assistance in development work. I see Anthropic leaning toward specializing in app development and task completion via MCP. For instance, there's a rumor that Apple at WWDC 2025 is partnering with Anthropic for AI coding assistant.
There’s a lot of value packed in this piece. Honestly, I see at least three standalone articles here:
One just on linearMCP. As I read it, I kept thinking, wait, what is this exactly? Did you invent it? How do all the pieces connect? Definitely deserves a deep dive.
Another could focus on the Claude integration, especially image manipulation, Slack workflows, etc. That section really caught my eye. I’ve struggled with image processing myself, so would love a technical breakdown there.
Third, your take on ClaudeCode vs. Cursor. You clearly prefer ClaudeCode, whereas I lean toward Cursor (especially in YOLO mode, it usually “just works” for me). Would be great to hear more on why you prefer ClaudeCode, what the tipping points are.
Thanks. I actually have an in progress article about my favorite mcp servers. Linear is our task management system (it’s like jira)
And I also have one trying to compare Claude code and GitHub copilot, not quite the same. But will be finishing those in the coming weeks. Thanks again for reading!!
Very clear argument for working in LLMs. I’m also surprised that it’s handling end to end (for some tasks). Wouldn’t have expected this outcome from my POV two months ago, but who am I to doubt 😅 can’t wait to try
Compared to other AI models from companies like OpenAI and Google, I keep returning to Anthropic's Claude model for assistance in development work. I see Anthropic leaning toward specializing in app development and task completion via MCP. For instance, there's a rumor that Apple at WWDC 2025 is partnering with Anthropic for AI coding assistant.
I think you’re right with Anthropocene leading for developer assistance.
Also with apple,
I hope so because Siri needs HELP!
Thanks a ton for introducing Claude to me. It’s been a real game changer!
can’t wait to discuss this next week!
Also completely agree on two things:
It’s easy to start too many AI-driven projects, and the context-switching becomes a killer.
With solid planning, AI coding can go far. But you need to know what you’re building.
Appreciate you putting this together!!
There’s a lot of value packed in this piece. Honestly, I see at least three standalone articles here:
One just on linearMCP. As I read it, I kept thinking, wait, what is this exactly? Did you invent it? How do all the pieces connect? Definitely deserves a deep dive.
Another could focus on the Claude integration, especially image manipulation, Slack workflows, etc. That section really caught my eye. I’ve struggled with image processing myself, so would love a technical breakdown there.
Third, your take on ClaudeCode vs. Cursor. You clearly prefer ClaudeCode, whereas I lean toward Cursor (especially in YOLO mode, it usually “just works” for me). Would be great to hear more on why you prefer ClaudeCode, what the tipping points are.
Thanks. I actually have an in progress article about my favorite mcp servers. Linear is our task management system (it’s like jira)
And I also have one trying to compare Claude code and GitHub copilot, not quite the same. But will be finishing those in the coming weeks. Thanks again for reading!!
Fantastic! Looking forward to reading them!
This is very intriguing, I will need to try claude on some new personal projects
Smartest dev I know!
LOL
Very clear argument for working in LLMs. I’m also surprised that it’s handling end to end (for some tasks). Wouldn’t have expected this outcome from my POV two months ago, but who am I to doubt 😅 can’t wait to try