Tempreon

Tempreon is a personal memory that AI assistants can read and write for you.

You store things once: facts about your work, decisions and the reasons behind them, the people you work with, your preferences for how things should be written and done, tasks, and files. Any assistant you connect to Tempreon can then use the same record, so what you stored while working in ChatGPT is available in Codex, and the reverse.

When you start a session, Tempreon returns your current focus, your active projects and the preferences you have taught. Search runs over your own stored records, not the public web. When you ask for work in your own voice, the assistant can get your voice, your working rules and the limits you have set from Tempreon. What you ask it to keep is saved through the same connection. In ChatGPT, Tempreon can also show a small card with what it is using for you right now and how to export your data.

You can also define your own agents and, when you ask, run one on a task in the background. Before an agent run starts, Tempreon returns an estimate and waits for your confirmation.

Who it is for: individuals who use more than one AI assistant and want them to share one memory that the person owns.

Limitations: Tempreon requires a Tempreon account. Its memory tools act only on the signed-in person's own records, never on other users' data, and do not search the web. An agent the person chooses to dispatch runs at an external model provider and can use the web.