Entoli (εντολή) is the Greek word for a command, both the one you give and the one the agent runs.
Entoli is a general-purpose open-source assistant designed to give you full-control. It uses Byte Repair technologies to securely check the code it generates in order to ensure it is safe and can be trusted. In the age of agentic-coding, LLMs can create code that has vulnerabilities (i.e. memory leaks, overflow/underflow, array bounds access failures, etc.). Byte Repair is designed to find these vulnerabilities and automatically fix them. Byte Repair automatically encodes the program as a mathematically constrained equation to find vulnerabilities.
Everywhere You Work
The same assistant runs on Linux desktop, Android, and the web, secure your code while your out living your life. Using source-control sync your work across-devices and make it work for you!
Client-Side By Design
Entoli calls the model endpoints you configure and runs the agent’s commands on your own machine. There is no server in the path, nothing to host, and no copy of your work anywhere you did not put it.
Most assistants are a service: your input goes to someone else’s machine, gets processed under someone else’s terms, and you take their word for what happened to it. Entoli inverts that. Your keys, your endpoints, your hardware. The only thing you install is the app.
Entoli is compatible with OpenAI style endpoints and Anthropic endpoints. For privacy, we recommend self-hosting or using Ollama Cloud.
Byte Repair Inside
One part of Entoli is the Byte Repair workspace for verification-backed coding: designed to seamlessly allow you to securely code as if you were doing any other agentic-coding.
Licence
AGPL-3.0, and independent of the institutional research programme behind the other products.