Welcome to the Hezo blog. This is where we'll share product news, release notes, and the occasional deep dive into how Hezo works under the hood — from agent isolation to budgets and long-term memory.
Why a blog
Hezo moves quickly, and a lot of what ships is easy to miss if you aren't watching the commit log. Release highlights land here first, in a form you can actually read: what changed, why it matters, and how to try it on your own hardware.
Expect three kinds of posts:
- Releases — what's new, with upgrade notes when they matter.
- Engineering — how the pieces work: orgs, projects, isolation, memory.
- Announcements — events, milestones, and things worth celebrating.
Writing a post
Posts are plain markdown files in the website repo. The filename becomes the URL slug, and a little frontmatter carries the metadata:
---
title: "Your post title"
date: "2026-07-03"
tags: ["release"]
image: "/news/your-post-image.png"
description: "One-sentence summary used for previews and social cards."
author: "Jane Doe" # optional byline
updated: "2026-08-01" # optional; set when you revise a post
faq: # optional; renders an FAQ section
- q: "A question people actually ask?"
a: "A brief, self-contained answer."
---Drop the file in data/news/, add the image to static/news/, and the
site build does the rest — the article gets its own page and a card on the
news index, newest first. Only title and date are required; the
optional fields switch on extras where they fit. The FAQ at the bottom of
this very post is rendered from its faq frontmatter — on the page for
readers, and as FAQPage structured data for search engines and AI
assistants. The full authoring guide lives in data/README.md.
That's it for now. If you're new to Hezo, the docs are the best place to start, and the project is open source on GitHub if you'd like to follow along or contribute. More soon.