Skip to main content
The compliance guardrail decides with a single Claude judge reading your content against one policy. A policy is the org’s compliance context plus a set of rules; exactly one policy is active per org. The judge returns a list of violations — one confirmed violation means fail, and clean content means an empty list, not a report of everything that passed. The policy you write is the prompt the judge runs on, so the quality of the rules is the quality of the decision.

A policy = org context + rules

A policy is your compliance context plus a set of rules. One active policy per org applies. Each rule has a title and a guidance string, plus optional good/bad example pairs. Two things to know:
  • There is no severity field. A flag is a flag; any confirmed violation blocks. You don’t rank rules by severity; you write them so each one only fires when it should.
  • The guidance must carry both halves. Every rule states what to flag and the carve-out — “flag X, but do NOT flag Y.” A rule with only the flag half over-flags; the carve-out is what keeps the judge on the line you actually mean.

The policy is the prompt

There’s no hidden rulebook. The judge’s prompt is assembled entirely from the policy: a system role (“an impartial legal/compliance reviewer for marketing content in a regulated industry”), then your ## Organization context, then ## Rules (each rule with its good/bad examples), then the ## Content to review. Nothing else feeds the decision — which is why calibrating the policy is how you calibrate the judge. (See Calibrate the guardrail.)

What it returns, and the verdict rule

The judge returns a list of violations, each with the rule it breaches, the offending excerpt, an explanation, and a suggested fix. The verdict is simple: status is fail if there’s at least one confirmed violation; otherwise it passes with an empty violation list. Clean content comes back empty: the guardrail gates content, so it doesn’t hand back a list of everything that was fine.
An expanded guardrail run showing several violations, each with a rule name, the flagged excerpt in quotes, an explanation, and a Fix suggestion.
Related: Set up the compliance guardrail · Calibrate the guardrail · Refiner vs guardrail · Automate compliance checks