Blue Jacket ConsultancyBlue Jacket Consultancy
Operating Log
Operating Log

June 8, 2026 · AI Systems · Agent Architecture

A Constitution Won’t Save You

You can write a full governing document for your AI agents, share it across the entire stack, and they will still drift.

The Finding

A recent Anthropic talk made the case from one direction: cut the bulk of the system prompt and move the logic into skills — progressive disclosure instead of one long prompt. I had already learned it from the other direction.

About six hours before that talk crossed my feed, I spent the better part of a day pressure-testing my own agent architecture — specifically the orchestration layer that runs my planning agent. I was trying to answer a simple question: how much of the system prompt is actually doing anything?

The answer was uncomfortable. I had been building a bottleneck and calling it governance.

Here is what I found. You can write a full constitution for your agents. You can make it shareable between the orchestration and execution layers. You can enumerate every rule, every constraint, every fallback. The agents will read it, acknowledge it, and then drift anyway — assuming authority where they should not, filling gaps in ways you did not intend.

The rule list doesn’t hold. The long prompt doesn’t hold.

What Governs Instead

The problem isn’t that the agents are disobedient. The problem is that you are asking a single context window to hold too much authority at once. When context grows large and ambiguous, the model fills the gaps. That filling is the drift.

The fix is architectural, not textual. Make the authority hierarchy a property of the system design, not a paragraph in the prompt. Skills load on demand. Subagents get narrow scope. The orchestration layer does not carry the full constitution — it carries a pointer to what is needed right now.

This is defense before offense. Substrate before agents. You do not govern by writing more rules into the prompt. You govern by shrinking what any single agent is authorized to know and do at any given moment.

What Transfers

I have been saying a version of this about revenue systems for years: you build the governance model before you scale the motion. The same principle applies to an AI stack. The teams that figure this out first will not win because they have better models. They will win because they built an architecture where the models cannot make the expensive mistakes.

The constitution is a document. The architecture is the actual governance.

Steady on.

Subscribe to the Operating Log

Field notes on building governed, production AI systems — new issues to your inbox.

Subscribe on Substack →

© 2026 Blue Jacket Businesses LLC, d/b/a Blue Jacket Consultancy. All rights reserved. May be shared in unmodified form with attribution. No use for machine-learning training, fine-tuning, or dataset construction without written permission.