The problem
A gate is only as good as the rules it holds.
Reviewers need confidence that the policy applied to a change is the one the team agreed on. When a change can reach into the very rules that grade it, that confidence slips — and a gate the team stops trusting is a gate the team routes around.
Keeping the standard fixed, review after review, is what makes the verdict something a team can lean on.
How it works
Rules evaluated from a trusted base.
When a change touches the gate's own policy, Cerebe evaluates that policy from the trusted base already in force. Every change is judged against the rules the team put in place, so the goalposts stay where they were set and the gate stays honest across every review.
The bind is part of the record: the decision is stored with the change, so anyone can see exactly which rules judged it. The same trusted baseline governs the workstation and the hosted check.
- The policy in force is evaluated from the trusted base already agreed on
- Every change is judged against those rules, keeping the standard consistent
- The same trusted baseline governs the workstation and the hosted check
- The decision is recorded with the change, so the bind is auditable
A gate that holds its standard.
Install the CLI for a local verdict, or turn on the GitHub App to gate every pull request.