Skip to main content

Independent Authorization for Agent Change

TrustRelay observes every agent-authored pull request, assembles verifiable evidence, applies deterministic policy, and issues a signed Agent Change Passport — before anything merges.

The Challenge

Agent-authored code needs independent oversight

AI coding agents are writing more code than ever, but existing review processes weren't designed for autonomous software contribution. TrustRelay fills the gap between agent productivity and safe deployment.

No Attribution

Agent-authored changes lack clear attribution and human sponsorship. Who is responsible when an agent makes a mistake?

Inconsistent Evidence

Without a standardized evidence pipeline, reviewers cannot reliably assess the safety and provenance of agent contributions.

Manual Review Bottleneck

Requiring human review for every agent change negates the productivity gains of AI coding assistants.

The Solution

Independent authorization that earns trust

TrustRelay is a deterministic authorization layer that observes every agent-authored pull request, assembles a verifiable evidence snapshot, and applies your policy before anything merges.

Deterministic Policy

Every authorization decision is reproducible from exact versioned inputs. No AI makes or modifies a decision — only typed assertions with source qualification.

Signed Evidence

Every change produces a KMS-signed Agent Change Passport stored in an immutable S3 ledger. Passports are independently verifiable by any party.

Required Enforcement

TrustRelay publishes a required GitHub check. If the check fails, the PR cannot merge — enforcement is architectural, not advisory.

Customer Control

You define the policy. You control the recovery path. TrustRelay provides the deterministic framework for your authorization rules.

How It Works

From pull request to decision in four steps

TrustRelay integrates at the merge boundary, providing independent authorization without disrupting developer workflows.

01

Observe

TrustRelay monitors GitHub pull requests where AI coding agents have contributed code changes.

02

Analyze

Agent attribution and human sponsorship are established. A versioned evidence snapshot is assembled from repository data, CI results, and policy context.

03

Decide

Deterministic policy evaluates the evidence against your configured rules. The result is APPROVE, BLOCK, or ESCALATE — each with a stable reason code.

04

Enforce

A signed Agent Change Passport is issued to the immutable ledger. A GitHub required check reports the outcome. Overrides and production outcomes are preserved.

Ready to delegate merge authority safely?

TrustRelay is the independent authorization layer your platform team needs to adopt AI coding agents at scale.