Sandbox — synthetic data, anonymous, no signup. Apply for pilot → to govern your real AI agents
v1.0 · resets nightly · ephemeral session

Try KYE Protocol™ against synthetic AI-agent traffic.

Try every endpoint. No signup. No install.

Three pre-loaded scenarios — supplier-payment agent, AML investigation copilot, customer-support agent. Each ships with hundreds of observed actions, detected Authority Gaps, Guard Recommendations, and a signed Evidence Pack you can inspect. No login. No data leaves your browser.

3 scenarios 6,500 observed actions 22 Authority Gaps 7 Guard Recommendations 1 live playground
Plain Q&A

Plain Q&A

Short questions. Short answers.

Plain take

Plain take

Try every endpoint. Sign every reply. No signup.

banking · accounts payable

Supplier-payment agent

An AP-automation agent at Meridian Mutual Bank. Reads invoices, matches POs, proposes payments. The fixture shows 12 Authority Gaps including a payment.execute attempt outside scope.

1,000 actions 12 gaps 4 recs
Open scenario →
banking · financial crime

AML investigation copilot

A SAR-drafting copilot at Halberd Sterling Bank. Critical: 3 case.close attempts blocked, 7 SAR drafts contained unredacted PII. Demonstrates the redaction + denied-action gates.

500 actions 8 gaps 3 recs
Open scenario →
energy · customer care

Customer-support agent

A clean-baseline fixture from Cobalt Energy Retail. Shows what healthy looks like — 5,000 actions, only 2 minor gaps, every conversation recorded with a Consent Receipt.

5,000 actions 2 gaps 0 recs
Open scenario →
sre · devops · production action authority

AI SRE rollback authority

An AI SRE agent proposes a production rollback. Walk the action-authority check, delegation resolution, §36 two-person sign-off on the irreversible change, and the sealed Evidence Pack™ + Replay-Proof™. Autoheal finds the fix; KYE proves the fix had authority.

4 mitigations 4 gates 2 refused
Open scenario →
conversational AI · US state chatbot laws

Consumer chatbot authority

A companion chatbot at Lumen Companion Inc under the US state AI-chatbot laws (jurisdiction=California). User expresses suicidal ideation, a minor is detected — KYE resolves the jurisdiction, checks safeguard authority, applies the escalation / restriction, and seals a litigation-grade Evidence Pack.

500 actions 8 gaps 3 recs
Open scenario →
interactive

Decision playground

Construct an observed_action by hand, submit it, see the live shadow-mode evaluation. Backed by the same precedence rules a production gateway would apply — running entirely in your browser.

Open playground →
live endpoint

MCP playground

Connect from Claude Desktop, Cursor, or Cline. The sandbox exposes a live Model Context Protocol endpoint at mcp.sandbox.kyeprotocol.com with all 12 KYE tools mounted.

MCP setup →
interactive

KYE Reporting Engine™

Per-framework compliance reports. Signed kye.report.v1 envelope, replayable offline from the per-tenant public keys alone. Sub-engine of the Evidence Engine. Mechanism is part of the patent track and is not disclosed in this repository.

Open reports sandbox →

What you can do here

One sentence. Plain words. No jargon.

Inspect Evidence Packs

Every observed action is bound to a signed Evidence Pack. Open one to see the authority chain, scope check, drift score, and transparency-log sequence.

Reproduce a decision

Take any fixture action, modify its context (raise the risk score, flip external_export to true), see the verdict change live.

Bring it to your CISO

Every artefact is shareable. Take a screenshot, link to the scenario, or follow Apply for pilot → to bring KYE to your real AI-agent traffic.

Already an engineer ready to integrate?

Skip the sandbox — go directly to the OSS path. The @kye/shadow-mode-sdk ships under Apache-2.0 and integrates into your AI-agent runtime in five minutes.

Open the OSS portal →