Governed execution
What is governed execution?
The category defined from first principles, and eight questions worth asking of any platform that claims to govern execution.
Resources
Understanding governed execution.
Where to read about governed execution, AI agent operations, verification, and operational evidence — organized by the question you are trying to answer.
Start with the question you have.
Governed execution is an emerging category, and the library is small. Everything listed here exists today; longer essays will be added when they are published rather than announced ahead of time.
Execution governance
Authorization is not execution
Why being allowed to ask — through an Application, a credential, or a tool — is a different decision from being allowed to run, shown step by step.
AI agent operations
When agents inherit shell access
The difference between handing an AI agent credentials and handing it a governed request path, with the scenarios it applies to.
Verification
Verified operational outcomes
Why a provider or an agent reporting success is an execution result, and what it takes to turn that into a verified outcome.
Operational evidence
Audit logs versus operational outcomes
The questions asked three weeks after a change, and what an evidence record can and cannot prove.
Platform engineering
One estate, multiple operational purposes
How purpose-scoped Applications give teams their own surface over shared infrastructure without sharing authority.
Open source
Resource Shell(opens in a new tab)
The open-source foundation for resource-oriented execution, and the origin of resh's resource model.
Program
Design partner program
Test governed execution against one consequential workflow of your own, in a controlled, non-production evaluation.
Ahead of the writing
The conversations are further along than the essays.
If you would rather not wait for the published version, talk to us directly. Design partner conversations cover the same ground in more depth, against your environment rather than a generic example — and what we learn in them is what ends up getting written.