Tips, behind-the-scenes, and news on personal AI agents.
Chasing a late payment costs the average affected UK business 86 hours a year (London Economics/DBT, 2025). The full amicable recovery process for England and Wales, with copyable templates for each stage, statutory interest and compensation explained, and what changes under the incoming Small Business Protections Bill.
Read the article →ChatGPT has memory now, and Projects to structure it. But it's still a tool you open to talk to — not an agent that runs continuously and writes to you first. Here's what actually changes, without caricaturing ChatGPT.
Read the article →CrewAI is a Python framework for developers, not a product you use directly. Comparing the two feature-by-feature doesn't really make sense — except to figure out which one matches what you actually know how to do.
Read the article →Fyxer does one thing — inbox and meetings — and does it in depth for teams. Hermes Agent covers a wider scope for one person. The right choice depends on what you actually want to delegate.
Read the article →Lindy is a closed SaaS platform built for teams, with credits that run out every month. Hermes Agent is open-source software built for one person. Here's what actually differs between the two, without flattering either.
Read the article →Manus has its own e-mail address and a Telegram agent, same as Hermes Agent. The difference isn't the channels — it's what happens between the tasks you give it.
Read the article →n8n is a node-based workflow automation tool, not a conversational agent. And its license explicitly forbids what we do with Hermes Agent. Two good reasons not to put them on the same footing.
Read the article →Yes, the software is free and open source. What isn't free: the server, the language model calls, and the hours you'll spend maintaining it. Here's the real total, with the math shown so you can check it yourself.
Read the article →Where does the data you hand to an AI assistant actually go? Named subprocessors, what flows through and what never does, explained plainly.
Read the article →Free, pay-per-use, or managed plan: here's what an AI assistant actually costs in 2026, with the pricing traps worth knowing before you choose.
Read the article →Already installed Hermes Agent yourself? Here's an honest look at what transfers to a managed service, what starts from zero, and how long it really takes.
Read the article →AI agent, chatbot, ChatGPT: three terms often confused. This guide details what really sets them apart, with a comparison table, to know which one you need.
Read the article →Quotes, unpaid invoice reminders, email sorting: what AI actually automates for a craft business or small company today, and the tasks it should never do alone.
Read the article →Invoices, reminders, morning briefings: here's what an AI assistant actually does for a freelancer, told through a typical day rather than a feature list.
Read the article →Hermes Agent is an open-source software (MIT license) published by Nous Research, a personal AI agent that runs continuously on a server. This guide explains what it can do, what it can't do, its real cost, and who it's for.
Read the article →Hermes Agent and OpenClaw are both open source and MIT-licensed, but with different publishers, architectures, and channel ecosystems. Here's what actually changes between the two, and how to migrate an existing OpenClaw configuration.
Read the article →Nous Research says a $5 VPS is enough, which is technically true but optimistic in real-world use. Here's the RAM you actually need, current VPS plans to compare, and why the server cost is only part of the bill.
Read the article →Native Windows support for Hermes Agent is experimental: WSL2 is the real target. This guide covers what comes before the installation itself — enabling WSL2, checking the version, tuning allocated memory — so you don't add one more blocker.
Read the article →Installing Hermes Agent actually takes 3 to 4 hours, not 5 minutes. Tested on August 1, 2026 on WSL2: a postinstall script stuck for 30 minutes, a build that fails under Node 20, a gateway failing silently. Here are the exact 6 pitfalls we hit, and how to avoid them one by one.
Read the article →