How a Habr author created an AI assistant on Hermes Agent for family and support in Telegram

A Habr author shared personal experience using the Hermes Agent IA assistant in Telegram — from a joint pet project with his wife to applying the agent to other tasks, including customer support.

How the project started

According to the author, the idea came from a small joint project with his wife: she handles the creative part, he handles the code. After building a project skeleton in Codex, the author created a group chat with three people — himself, his wife, and Hermes Agent — shared with the agent access and project context. As a result, his wife got the ability to independently add features and move interface elements by sending voice messages directly to the shared chat.

• Project skeleton made in Codex
• A Telegram group chat created with three members: the author, his wife, and Hermes Agent
• Wife makes interface corrections through voice messages without writing any code
• According to the author, the experiment worked so well that he expanded the use of Hermes Agent to other scenarios

Where else Hermes Agent proved useful

The author notes that the successful experience with the family project prompted him to use Hermes Agent in other scenarios — including customer support and personal pet projects. In the article, he promises to explain what Hermes Agent is, how exactly he applies it, and how readers can set up a similar assistant for themselves.

Why this is interesting for developers

The key detail of the case is that access to the agent was given not only to the author-developer himself, but also to a person without technical background. The author's wife interacts with Hermes Agent by voice, not through code or console, and the agent independently translates voice instructions into specific interface changes and functionality improvements of the project. It is precisely this scenario of collaboration between a developer and a non-technical person through the same agent that the author considers the most valuable result of the experiment.

At the same time, it's important that the entire process happens without intermediate retellings: the wife doesn't explain to her husband in words what needs to be fixed in the interface, and the husband doesn't translate her wishes into a task for himself — both turn directly to the same agent with shared access to the code and project context, and the agent itself delivers the fix to completion.

What's behind the format of "group chat with an agent"

The format described by the author is not a separate application or bot with a fixed set of commands, but a full participant in a shared Telegram chat on an equal basis with people. Both people see the same conversation with the agent, both can intervene in a task at any moment, and the agent itself responds to messages from both participants in the shared project context. This differs from the classical scheme where each developer has their own separate access to the assistant and their own conversation with it — here the entire context is shared and synchronous for the entire team of two people.

What this means

The case demonstrates how AI agents like Hermes Agent lower the entry barrier into development: a person without technical background can directly influence the product through voice without writing code, if the agent is given access to the right context and tools. For small teams and family pet projects, this means that the role of "second developer" can be taken on not by a person but by an agent that both participants trust with the same set of permissions. The author emphasizes that precisely this effect — and not a separate feature or line of code — became for him the main discovery of the entire experiment.