March 25, 2026
Merge bots or chaos gremlins?
Show HN: Optio – Orchestrate AI coding agents in K8s to go from ticket to PR
From to-do to merged code—no humans? Commenters split between brave and bonkers
TLDR: Optio promises a bot that takes a task and automatically ships merged code by fixing tests and responding to reviews. Commenters are fascinated but skeptical, debating safety, isolation for multiple customers, and real-world proof—while joking that misaligned README diagrams might be the canary for misaligned code.
Meet Optio: the launch that claims it can take a to-do ticket and turn it into merged code by itself. In plain English, it’s a bot that reads your task, writes the code, opens the pull request (the proposal to change your app), waits for tests to pass, fixes what breaks, responds to reviews, and merges. The pitch: a self-driving coding pipeline with a “feedback loop” that keeps the bot working until the job is done.
The crowd reaction? Equal parts applause and side-eye. One early congrats quickly turned into a security grilling as MrDarcy asked whether there’s “any sandbox isolation from the k8s platform layer?”—translation: if this runs on Kubernetes (the server system), can it safely handle multiple customers without cross-contamination. Reliability anxiety lit up next: antihero worried, “what stops it making total garbage that wrecks your codebase?” Others demanded receipts, with conception asking for the most complicated thing Optio has actually shipped. Then came the vibe check: knollimar joked the post “sounds like satire,” suggesting the promise is so bold it reads like an AI wrote its own hype.
Meanwhile, the peanut gallery found a new villain: wonky ASCII art in the README. hmokiguess couldn’t unsee the misaligned diagram—because if the demo can’t align columns, can it align commits? That’s the tone of the thread: big vision, big questions, and a meme-ready divide between “ship it” and “shipwreck.”
Key Points
- •Optio automates the path from task (web UI, GitHub Issues, or Linear) to merged pull request using AI coding agents.
- •A continuous feedback loop resumes agents on CI failures, merge conflicts, and review feedback until PRs are approved and passing.
- •It runs a pod-per-repo architecture on Kubernetes with persistent volumes, multi-pod scaling, health monitoring, and idle cleanup.
- •Optio can launch a separate code review agent and block merging until the review completes, using configurable, possibly cheaper models.
- •Per-repo configuration includes model selection (Claude variants), container images or custom Dockerfiles, Handlebars prompt templates, concurrency limits, and setup scripts.