Execution Governance Developer Docs

Documentation

Introduction

11/11 Execution Governance is sovereign runtime infrastructure that decides whether an AI, agent, or autonomous system is allowed to act, before the action reaches the real world. Authorization is evaluated before execution, enforced during runtime, and proven after the fact.

Quickstart

The 11/11 Execution Governance SDK is open source under Apache-2.0. Install it from npm and run it entirely on your machine, no API key, no network, no telemetry:

npm install @11ai/execution-governance

Create a gate from your policy and wrap any function call. The gate evaluates before execution and runs the action only on allow, throwing DeniedError on deny, with a signed receipt either way:

import { createGate } from "@11ai/execution-governance";

const gate = createGate({ policy: "./eg-policy.yaml", signingKey });

const result = await gate.govern(
  { sessionId: "s1", tool: "http.post", args: { url, body } },
  () => fetch(url, { method: "POST", body })
);

Authentication

The open-source SDK requires no API key, no account, and no network, it runs entirely on your machine. API keys apply only to the optional hosted control plane (control.11aiblockchain.com), which authenticates requests with a key issued through onboarding; store it as a secret, never commit it, and rotate on suspected exposure.

Execution Governance

Execution Governance is the pre-execution authorization layer. Every action is bound to an identity (subject), an operation (action), a target (resource), and an environment (context), then evaluated against the active policy set before it is permitted to run. The result is a governance decision and, when permitted, an authorization artifact that gates downstream execution. Internal policy-engine logic is not exposed.

Governed Execution

Governed Execution is runtime enforcement. An action proceeds only when a valid authorization artifact is present; without it, execution is stopped. Enforcement is fail-closed: the default outcome in the absence of explicit authorization is denial. It is fail closed by design and enforced in line, subject to independent verification.

EA-11 Execution Arithmetic

EA-11 produces a deterministic, Merkle-rooted evidence state for every governed execution. Each decision is reduced to a set of component hashes that can be independently verified. The underlying arithmetic and internal math are proprietary and are not published.

Fail-Closed Governance

Fail-closed means the default is deny. If authorization is missing, invalid, or cannot be evaluated, the action does not proceed.

SDK Overview

The Execution Governance SDK is a local, typed library for governed execution, signed-receipt generation, and offline receipt verification, no network or control plane required. It can optionally integrate with the hosted control plane for multi-tenant policy management and a public proof ledger.

JavaScript SDK

Install and use the JavaScript / TypeScript client.

import { verifyReceiptFile } from "@11ai/execution-governance";

const report = await verifyReceiptFile("./eg-receipts.jsonl", publicKey);
// report.ok is true when every receipt signature and the hash chain verify

Or from the command line, with no code: npx eg-verify --receipts ./eg-receipts.jsonl --pubkey <public key>.

API Overview

Public infrastructure surfaces. Endpoints beyond those listed are described as conceptual interface documentation only.

Authorization API

The authorization flow evaluates a request and returns a governance decision and, when permitted, a proof artifact.

Request → Policy Evaluation → Governance Decision → Execution Permission → Proof Artifact

Proof API

Each governed execution can be verified against the public proof surface. Both APPROVED and DENIED decisions are signed and recorded. Human-readable proof is available at the public Proof Viewer.

Health API

The public health endpoint reports infrastructure status.

GET https://control.11aiblockchain.com/health

Financial Profile

Governed execution for financial infrastructure. High-value transactions and settlement actions are authorized before they execute, with per-tenant isolation and signed, verifiable evidence for every decision. Direct payment-rail and stablecoin integrations are on the roadmap.

Healthcare Profile

Governed execution for clinical and healthcare AI. HIPAA-aligned decision boundaries are enforced before patient-impacting actions occur, with append-only, tamper-evident audit lineage and fail-closed operational controls.

Defense Profile

Governed execution for sovereign, regulated, and mission-critical environments. Fail-closed authorization gates every action with tenant-scoped isolation, cryptographic execution controls, and independently verifiable evidence for audit and oversight.

Conformance Overview

Conformance profiles describe requirement sets for governed execution across sectors, mapped to recognized frameworks where applicable.

Resources

Public surfaces, the main site, and contact.