2026년 5월 31일의 Show HN

Hi HN, I’m working on Ouijit.

It’s a project and task-based terminal session manager that provides a few basic but useful tools for agent workflows:

- Terminal sessions in Ouijit have access to the ouijit CLI, and supported agents (Claude, Codex, Pi) can work with it out of the box to manage tasks and customize a personal development workflow

- Tasks live on a kanban board that supports hooks for task lifecycle events (eg. ‘Run this script when a task moves to ‘in progress’)

I’ve found this simple combination to be very expressive and flexible for adapting to changing workflows.

I made the V1 a couple months ago for fun, and have kept at it since a friend shared they had logged an 8 hour work session in it. Along the way I’ve baked in lots of what I believe are table-stakes for this type of tool, like task isolation via Git worktrees, agent working/idle status with sound and notifications, diff/markdown plan/URL previews, and support for VM sandboxing using Lima.

It’s free and open source with no login or telemetry, so feedback is highly appreciated.

Github: https://github.com/ouijit/ouijit Website: https://ouijit.com

I made my first macOS utility app that ships with a bundled Gemma 4 model, specifically the Gemma E4B one. It made my app DMG have 5.3 GB in size, but I think it is a small size for the power that this free local model can provide.

It runs fine on CPU, but can also run on Apple Silicon GPU, although I did not notice any performance improvements with GPU (tested on a M5 chip).

I think these local lightweight and multimodal models will open multiple possibilities for new software tools where privacy is essential.

I'm building Deliberate in response to the PocketOS incident in April and the DataTalks.Club incident in February. Jer Crane could see the API call that deleted his prod volume. He couldn't see that the agent had other options and picked the worst one. Trace tools show you what ran. I wanted something that shows you what the agent was choosing between. Early access for teams running LangGraph or OpenAI Agents in prod.

based on a conversation I had with Ramana Kumar in 2016.

Hello HN, Cordium is a general-purpose sandbox platform built on Kubernetes and Octelium, may main work https://github.com/octelium/octelium, that can be used for various use cases, including coding for developers with VSCode, Zed, etc. (i.e. self-hosted GitHub Codespaces alternative), AI agent tasks (i.e. FOSS alternative to AI sandbox products such as E2B, Daytona, etc.), CI/CD workloads (e.g. building and publishing Docker images etc.), and more importantly for secretless remote access to infrastructure for devs and automated workloads.

The main _differentiator_ here, compared to other dev environments and sandbox platforms, is that Cordium automatically provides identity-based, secretless secure access to resources/infrastructure (e.g. APIs, SSH, databases, k8s, etc.) without having to inject credentials (e.g. API keys, SSH private keys, database passwords, etc.) into the sandbox where the upstream credential is held by the identity-aware proxy of the Octelium-protected resource outside the reach of the sandbox. You can simply think of it as a sandbox + ZTNA/remote-access-VPN baked-in where access to infrastructure is based on identity and policy-as-code rather than credentials.

Cordium is a purely FOSS project under Apache 2.0 that's meant for self-hosting and there are no plans for a pro/SaaS/cloud version. The development of the project started back in 2022 and it is already being used by a few organizations that use Octelium since last year. Happy to answer any questions.

This is my little app. I was sick of copying images from Pinterest and doing the posterising in photoshop when I was doing value studies as painting practice so I made a site which does the entire workflow in one bit. Happy painting!

AI Ops Platform (Security, Optimization, Observability, Evaluations) Free for early signups and reviews