Hacker News

It seems that using Cursor has become really repetitive. My workflow typically goes like this: - Get a task to accomplish on a codebase - Research the web and codebase to create a plan - Send the entire plan to cursor and sit back for it to finish

This flow works really well with a high level of accuracy for the most part and it's becoming obvious that the margin between where it is today and "perfect" is quickly closing.

Because of this flow, I feel that I can replicate it through an SDK that uses a AI cli coder tool with low level control.

In this SDK, you can also use the best coder in the world per: https://aider.chat/docs/leaderboards/

As shown here:

```python from cloudcode import Local import os

def main(): # Use the current directory cwd = os.getcwd() example_file = "example.py"

# Initialize the Aider SDK in architect mode agent = Local( working_dir=cwd, model="o3", # Main (planner) model editor_model="gpt-4.1", # Editor model for implementing changes architect_mode=True, use_git=False, api_key=os.getenv("CLOUDCODE_API_KEY") ) # Create or overwrite a simple Python file to modify using the SDK agent.create_file( example_file, """def add(a, b): return a + b """ )

# Run a coding task using the two-model workflow result = agent.code( prompt="make this function super cool. just make the math functions better", editable_files=[example_file] ) # Print the results print("\nTask completed!") print(f"Success: {result['success']}") print("\nChanges made:") print(result["diff"]) # Display cost information print("\nTask completed!") if __name__ == "__main__": main()

```

Given that you can mimic your workflows from cursor and claude code into this sdk, you shouldn't have to sit around on your computer waiting for the long running task to finish.

That's why I built in a Sandbox class that mounts the ai coder into a remote sandbox directory instead of a local folder so you can deploy an ai coder to the cloud, enabling use cases such as coding from your phone and the browser.

I have already found so many low-handing fruit use cases for this sdk to automate prompting, information gathering, and auto-docs as shown in our github https://github.com/LMSystems-ai/cloud-coding

But I want to take it a step further.

You can get started for free today with this sdk and see our favorable pricing model here: https://docs.cloudcoding.ai/pricing

Comments URL: https://news.ycombinator.com/item?id=43761446

Points: 1

# Comments: 0

Article URL: https://make.wordpress.org/project/2025/04/16/a-new-cadence-for-wordpress-core/

Comments URL: https://news.ycombinator.com/item?id=43761436

Points: 1

# Comments: 0

Article URL: https://ssh-keys.arent-real.com

Comments URL: https://news.ycombinator.com/item?id=43761423

Points: 2

# Comments: 2

I have an idea for optimizations using a dedicated backend for pytorch. It only applies to eager mode execution. From my understanding, eager mode is used for development and debugging. Are there real production workloads that work in that way ? I assume people optimize by compiling graphs using JIT but then again, people are lazy...

Comments URL: https://news.ycombinator.com/item?id=43761400

Points: 1

# Comments: 0

Article URL: https://github.com/jbyuki/instant.nvim

Comments URL: https://news.ycombinator.com/item?id=43761395

Points: 1

# Comments: 0

Article URL: https://arxiv.org/abs/2402.14531

Comments URL: https://news.ycombinator.com/item?id=43761387

Points: 3

# Comments: 0

Article URL: https://cartai.app/

Comments URL: https://news.ycombinator.com/item?id=43761382

Points: 1

# Comments: 0

Article URL: https://restofworld.org/2025/skype-shutting-down/

Comments URL: https://news.ycombinator.com/item?id=43761373

Points: 2

# Comments: 0

Article URL: https://www.datacouncil.ai/talks25/look-ma-no-data-warehouse

Comments URL: https://news.ycombinator.com/item?id=43761351

Points: 2

# Comments: 0

Article URL: https://www.baesystems.com/en/article/major-breakthroughs-in-uk-munitions-production

Comments URL: https://news.ycombinator.com/item?id=43761348

Points: 1

# Comments: 0

Article URL: https://www.jalopnik.com/dune-was-originally-published-by-a-car-repair-manual-co-1847940372/

Comments URL: https://news.ycombinator.com/item?id=43761345

Points: 1

# Comments: 0

1、fastapi_mcp, open source, a simple way to convert：https://github.com/tadata-org/fastapi_mcp

2、Higress, open source, with a practical MCP market, powered by AI gateway：https://mcp.higress.ai/

3、RapidMPC: SaaS, no code, Unlimited MCP Servers, $12/month：https://rapid-mcp.com/

Comments URL: https://news.ycombinator.com/item?id=43761328

Points: 1

# Comments: 0

Article URL: https://www.nature.com/articles/d41586-025-01216-7

Comments URL: https://news.ycombinator.com/item?id=43761314

Points: 2

# Comments: 0

Article URL: https://github.com/tric/trick2025

Comments URL: https://news.ycombinator.com/item?id=43761083

Points: 1

# Comments: 0

Article URL: https://arstechnica.com/gadgets/2025/04/in-depth-with-windows-11-recall-and-what-microsoft-has-and-hasnt-fixed/

Comments URL: https://news.ycombinator.com/item?id=43761077

Points: 1

# Comments: 0

Article URL: https://www.cmu.edu/tepper/news/stories/2025/april/using-principles-of-swarm-intelligence-study-compared-platforms-that-allow-brainstorming-among-large-groups.html

Comments URL: https://news.ycombinator.com/item?id=43761050

Points: 1

# Comments: 0

Article URL: https://www.youtube.com/watch?v=JgYiAJgXFiw

Comments URL: https://news.ycombinator.com/item?id=43760998

Points: 1

# Comments: 1

Article URL: https://www.youtube.com/watch?v=-gbZunvvEds

Comments URL: https://news.ycombinator.com/item?id=43760992

Points: 1

# Comments: 1

Article URL: https://twitter.com/NTFabiano/status/1914289562144145622

Comments URL: https://news.ycombinator.com/item?id=43760991

Points: 2

# Comments: 0

Article URL: https://medium.com/@mustafaakin/my-biggest-entrepreneurial-regret-01bc20a211a4

Comments URL: https://news.ycombinator.com/item?id=43760982

Points: 2

# Comments: 0