How agents are making software provably secure
Agentic development is fuelling a resurgence in formal verification, enabling engineers to go beyond simply testing, to mathematically prove the robustness of software.
Conor Svensson
Technical founder building agentic infrastructure and tooling for the decentralised web.
Latest posts
From private chats to shared agents
Shared agents are changing how teams work with LLMs. But running them yourself can be a significant undertaking with security implications.
Why non-engineers ship unmaintainable code with AI agents
Coding agents are optimised to appease the operator, not to think critically. Without an operator who understands engineering, you get working demos that fall apart when someone tries to maintain them.
What we learned trying to give an agent our whole company
How we built text-native archives of our own operations so agents could work with them like codebases.