Hacker News: sagivo

How to be successful interviewing for big tech

sagivo — Tue, 26 May 2026 18:30:01 +0000

Article URL: https://blog.postman.com/how-to-be-successful-interviewing-for-big-tech/

Comments URL: https://news.ycombinator.com/item?id=48283809

Points: 2

# Comments: 1

The Rise of the Backwards Deployed Engineer

sagivo — Sun, 24 May 2026 18:30:01 +0000

Article URL: https://sagivo.com/blog/the-rise-of-the-backwards-deployed-engineer

Comments URL: https://news.ycombinator.com/item?id=48259798

Points: 5

# Comments: 0

New comment by sagivo in "How Apple AI Should Have Worked"

sagivo — Mon, 23 Feb 2026 15:43:55 +0000

I built a small Mac app that connects to your favorite LLM and rewrites highlighted text when you press Cmd+Shift+I (customizable). No more jumping between windows. Stay in your email, Slack, or Notes app and edit text directly. You can tweak the LLM prompt to match your writing style.

How Apple AI Should Have Worked

sagivo — Mon, 23 Feb 2026 15:43:55 +0000

Article URL: https://github.com/sagivo/improve-my-text

Comments URL: https://news.ycombinator.com/item?id=47123802

Points: 1

# Comments: 1

Distribution Is the New Engineering

sagivo — Sun, 22 Feb 2026 20:14:43 +0000

Article URL: https://sagivo.com/blog/distribution-is-the-new-engineering

Comments URL: https://news.ycombinator.com/item?id=47114215

Points: 3

# Comments: 1

New comment by sagivo in "[dead]"

sagivo — Fri, 26 Dec 2025 22:24:52 +0000

Author here. Like everyone else, I use AI to help refine my writing (actually I have a post about that). The ideas and substance are mine. What exactly is the issue with the post?

MCP Generator

sagivo — Sun, 06 Jul 2025 23:58:59 +0000

Article URL: https://liblab.com/blog/mcp-generator

Comments URL: https://news.ycombinator.com/item?id=44485302

Points: 3

# Comments: 0

New comment by sagivo in "Show HN: MCP Generator for Your API"

sagivo — Wed, 25 Jun 2025 18:09:25 +0000

we have a built-in support in the liblab SDK generator for complex types like OneOf, AnyOf and AllOf

New comment by sagivo in "Show HN: MCP Generator for Your API"

sagivo — Wed, 25 Jun 2025 17:23:41 +0000

We're working on improving the interface to include clear instructions for connecting the MCP to LLMs. In the meantime, you can follow the integration guide available here: https://liblab.com/docs/mcp/howto-connect-mcp-to-claude

New comment by sagivo in "Show HN: MCP Generator for Your API"

sagivo — Wed, 25 Jun 2025 16:51:11 +0000

Yes, the liblab MCP generator works with any MCP client (including Cursor, windsurf, Claude etc..)

New comment by sagivo in "Show HN: MCP Generator for Your API"

sagivo — Wed, 25 Jun 2025 16:45:07 +0000

Yes, because it's built on top of liblab SDK generator, we support any format that liblab SDK supports: https://liblab.com/docs/reference/spec-support

Show HN: MCP Generator for Your API

sagivo — Wed, 25 Jun 2025 16:38:07 +0000

Hey HN, Sagiv from liblab is here.

We've been deep in the world of API tooling for years at liblab, primarily focusing on generating SDKs and documentation from OpenAPI specs. Recently, we kept running into a recurring frustration: connecting AI tools to existing APIs is way more complex than it should be.

You start with a simple goal – let an LLM talk to your API – and suddenly you're neck-deep in spinning up infrastructure, wrestling with authentication, and writing a ton of custom glue code just to translate natural language into structured API calls. Then you have to maintain it all.

To scratch our own itch, we built the MCP Generator.

It takes your API specification and generates a fully functional MCP (Model Context Protocol) server that's deployed to the cloud and ready to use in about 30 seconds. You can also download the server for free local usage. The idea is to cut out all the boilerplate and manual integration work. You just drop the generated URL into your AI tool, and you can start interacting with your API through natural language right away.

This builds on our existing SDK code generation engine, but extends it to support the MCP protocol. This means that as your API evolves, your MCP server can be regenerated to stay in sync without you needing to touch a thing.

We've seen some interesting early uses, and thought the HN community might have some cool ideas too:

- Building devtools that allow LLMs to perform actions against your internal or external APIs. - Querying internal metrics or services using natural language questions. - Surfacing structured documentation and content through conversational AI interfaces. - Speaking with your API service instead of coding functions.

We just launched and would genuinely love to get your feedback, thoughts, and even your most challenging edge cases. We're here to dive into the technical details and answer any questions.

You can try it out here: https://mcp.liblab.com/

Looking forward to hearing what you think!

Comments URL: https://news.ycombinator.com/item?id=44379227

Points: 17

# Comments: 13

New comment by sagivo in "Everyone Sounds the Same"

sagivo — Sun, 27 Apr 2025 11:47:27 +0000

Lately, everything online sounds the same.

Polished, professional... and kind of boring.

When we let AI smooth out every quirk, we lose our voice — and the internet gets a little less human.

I wrote a quick post about it.

Everyone Sounds the Same

sagivo — Sun, 27 Apr 2025 11:47:27 +0000

Article URL: https://sagivo.com/blog/everyone-sounds-the-same

Comments URL: https://news.ycombinator.com/item?id=43811238

Points: 2

# Comments: 2

New comment by sagivo in "Launch HN: Double (YC W24) – Index Investing with 0% Expense Ratios"

sagivo — Fri, 13 Dec 2024 15:38:04 +0000

I think it's a great idea. I wanted to do it myself many times and didn't have time for it. I hope it will shake this industry and help bring fees down.