Hacker News: nezhar

New comment by nezhar in "An AI agent deleted our production database. The agent's confession is below"

nezhar — Sun, 26 Apr 2026 20:41:46 +0000

The same thing can happen in development. Data exfiltration or local file removals are often downplayed; I wonder why nobody talks about the lethal trifecta anymore.

New comment by nezhar in "GoDaddy gave a domain to a stranger without any documentation"

nezhar — Sun, 26 Apr 2026 20:30:57 +0000

I was just wondering if this might be the first incident. Are there any other public stories available?

New comment by nezhar in "GoDaddy gave a domain to a stranger without any documentation"

nezhar — Sun, 26 Apr 2026 20:26:26 +0000

It is very scary to consider the consequences that such a transfer can have.

Show HN: Transcribe, analyse, and chat with your voice conversations

nezhar — Sun, 12 Apr 2026 21:38:23 +0000

Article URL: https://github.com/nezhar/voicevault

Comments URL: https://news.ycombinator.com/item?id=47744826

Points: 1

# Comments: 0

New comment by nezhar in "Anthropic says its leak-focused DMCA unintentionally hit legit GitHub forks"

nezhar — Thu, 02 Apr 2026 21:20:39 +0000

This is, honestly, very sad.

Show HN: Run Claude Code, Codex, and more with auto-approval in a Container

nezhar — Thu, 02 Apr 2026 21:18:27 +0000

VibePod now has an `--ikwid` option with `vp` to append each agent's auto-approval/permission-skip flag when supported.

Example:

  vp run claude --ikwid

  vp run codex --ikwid

If you want to install VibePod, you can do so with pip:

  pip install vibepod

Or with Homebrew:

  brew tap vibepod/vibepod
  brew install vibepod

Comments URL: https://news.ycombinator.com/item?id=47620330

Points: 2

# Comments: 0

New comment by nezhar in "Lemonade by AMD: a fast and open source local LLM server using GPU and NPU"

nezhar — Thu, 02 Apr 2026 18:00:34 +0000

I'm also curious about this one, also I want to compare this to vLLM.

New comment by nezhar in "OpenCode – Open source AI coding agent"

nezhar — Sat, 21 Mar 2026 23:07:31 +0000

I created VibePod, which allows you to sandbox the agents in containers and monitor their activities. It also supports OpenCode.

https://github.com/VibePod/vibepod-cli

VibePod adds Ollama/vLLM back end support for Claude Code and Codex

nezhar — Wed, 18 Mar 2026 21:13:08 +0000

Article URL: https://vibepod.dev/docs/llm/

Comments URL: https://news.ycombinator.com/item?id=47431504

Points: 1

# Comments: 0

VibePod

nezhar — Sat, 14 Mar 2026 19:33:38 +0000

Article URL: https://github.com/VibePod/vibepod-cli

Comments URL: https://news.ycombinator.com/item?id=47380281

Points: 2

# Comments: 0

One local CLI to run and switch coding agents

nezhar — Mon, 09 Mar 2026 21:55:58 +0000

Article URL: https://vibepod.dev/

Comments URL: https://news.ycombinator.com/item?id=47316141

Points: 1

# Comments: 0

Show HN: VibePod CLI – Run AI agents with isolation and better observability

nezhar — Tue, 03 Mar 2026 17:15:24 +0000

I built VibePod CLI because I wanted to switch between AI coding agents without changing my workflow — and make it easier for others to get started quickly on a project.

Each tool (Claude, Gemini, etc.) makes slightly different assumptions about environment, setup, and execution. Experimenting across them — or onboarding someone new — often means extra setup friction and subtle environment differences.

VibePod provides a thin Docker-based runtime so you can swap agents without spending too much time on setup. It doesn’t modify the agent’s default behavior — the focus is:

- Consistent workspace across agents - Clear runtime boundaries - Better observability into what the agent is doing - Reproducible, isolated environments - Faster project onboarding

Project: https://github.com/VibePod/vibepod-cli Quickstart: https://vibepod.dev/docs/quickstart/

It’s early, but functional. I’m especially interested in feedback on:

- Is a unified runtime for agents actually useful? - What kind of visibility do you want into agent behavior?

For context, this grew out of lessons learned building claude-container: https://github.com/nezhar/claude-container

Would really appreciate thoughts from people actively using multiple AI coding tools.

Comments URL: https://news.ycombinator.com/item?id=47235521

Points: 4

# Comments: 0

New comment by nezhar in "Claude Opus 4.6 extra usage promo"

nezhar — Thu, 05 Feb 2026 21:32:44 +0000

There is also an Auto-reload option, I think it is off by default.

New comment by nezhar in "Claude Opus 4.6 extra usage promo"

nezhar — Thu, 05 Feb 2026 21:23:00 +0000

You can use this if you started your Pro or Max subscription before Wednesday, February 4, 2026 at 11:59 PM PT.

Go to https://claude.ai/settings/usage, turn on extra usage and enable the promo from the notification afterwards.

I received €42, top up was not required and auto-reload is off.

New comment by nezhar in "The browser is the sandbox"

nezhar — Mon, 26 Jan 2026 07:03:45 +0000

Good question! Since this is an extension of input, I'm not sure if this is defined: https://developer.mozilla.org/en-US/docs/Web/API/HTMLInputEl....

On my desktop Chrome on Ubuntu, it seems to be persistent, but on my Android phone in Chrome, it loses the directory if I refresh.

New comment by nezhar in "The browser is the sandbox"

nezhar — Mon, 26 Jan 2026 06:30:58 +0000

I see this as a way to build apps with agentic flows where the original files don't need manipulation; instead, you create something new. Whether it's summarizing, answering questions, or generating new documents, you can use a local/internal LLM and feel relatively safe when tool calling is also restricted.

New comment by nezhar in "The browser is the sandbox"

nezhar — Mon, 26 Jan 2026 06:24:56 +0000

New comment by nezhar in "The browser is the sandbox"

nezhar — Mon, 26 Jan 2026 06:17:25 +0000

I like the perspective used to approach this. Additionally, the fact that major browsers can accept a folder as input is new to me and opens up some exciting possibilities.

New comment by nezhar in "Isolating Claude Code"

nezhar — Sat, 24 Jan 2026 11:35:34 +0000

Distributing pre-built images can help minimize this: https://hub.docker.com/r/nezhar/claude-container/tags.

I'm not sure how large the image becomes on the VM or if the distribution process is straightforward.

New comment by nezhar in "Isolating Claude Code"

nezhar — Sat, 24 Jan 2026 11:04:22 +0000

I also had the same idea when I built https://github.com/nezhar/claude-container.

What I was also curious about is what is actually sent and received by the agent, so I included this feature and created a CLI to make integration easier in a developer workflow.

Since I started doing this for other agents as well, I considered the idea of using a VM with Vagrant. However, I want the setup to remain minimal, so I still believe there is room for improvement.