Gavriel Cohen on Why AI Agents Kill Restrictive Licenses
Gavriel Cohen explains why coding agents make restrictive licenses obsolete, how NanoCo monetizes, and his rule for filtering pull requests.
From zero to NanoClaw (Interview)
With Adam Stacoviak, Gavriel Cohen · Sunday, October 11, 2026
Gavriel Cohen, creator of NanoClaw and co-founder of NanoCo, sits down with Adam Stacoviak to discuss how a 40-hour weekend project evolved into a viral open-source agent framework and an enterprise company. They dive into open source licensing in the era of coding agents, container sandboxing, dual-database SQLite architecture, and why human workers must learn to become managers of hundreds of agents.
Key takeaways
- Restrictive software licenses fail when coding agents can ingest a codebase and translate it into a different programming language overnight. Read more →
- Asking an LLM to police its own boundaries fails; compliance rules belong in deterministic, non-AI code outside the model context window. Read more →
- Gavriel Cohen built NanoClaw as a 40-hour weekend project before turning the open-source agent framework into an enterprise company, NanoCo. Read more →
3 articles from this episode
Never Ask an AI Agent to Police Itself
Gavriel Cohen explains how NanoClaw uses Docker sandboxes and non-AI approval gates to secure enterprise agents.
Why You Must Spend 50% of Your Time Building Agent Environments
NanoClaw creator Gavriel Cohen explains why writing code matters less than designing feedback loops, linters, and context for autonomous agents.