Sovereign Compliance Active

Technical Evidence
for AI Regulatory Mapping

Sentinel provides the engineering infrastructure to discover, verify, and document AI system technicalities against regulatory patterns. From local AST discovery to immutable governance logs.

bash — 80x24

$ npx @radu_api/sentinel-scan ./manifest.json

Sentinel scans AI systems, verifies regulatory requirements, and generates technical evidence required for compliance audits.

Run your first audit View on GitHub

Open Source

NPM Package

GitHub Action

Why Sentinel exists

Regulatory transparency is now software infrastructure.

Engineering teams require technical evidence trails, just like they need unit tests, security scans, and build logs.

Sentinel maps regulatory patterns from the EU AI Act into deterministic engineering rules that assist in structuring technical documentation directly within development workflows.

The Sentinel Trinity

Triple-Layer Strategy

A unified technical framework to discover your AI footprint, verify it against regulatory patterns, and manage immutable evidence.

01. Scout

Autonomous discovery engine that crawls your repositories to identify AI signals, third-party LLM calls, and training data markers.

AI Signal Inventory
Automated Manifesting
AST Traversal

Explore Discovery

02. Engine

Deterministic WASM engine that verifies software technicalities against specific regulatory articles in real-time.

Local Deterministic Scan
Zero-Egress Processing
WASM Isolation

View Specifications

03. Hub

Immutable D1 ledger that provides technical evidence management, versioned snapshots, and automated Annex IV drafting.

Audit Integrity Ledger
Evidence Snapshots
Draft Synthesis

Inspect Ledger

Try Sentinel in 30 seconds

Install the Engine, run a local scan, and map technical signals to regulatory patterns in seconds.

Install

npx @radu_api/sentinel-scan

Pull the latest Sentinel engine directly via NPX. No global installation needed.

Run a scan

npx @radu_api/sentinel-scan .

Scan your local repository and surface compliance gaps, missing documentation, and actionable rule violations.

See the result

❌ Sentinel technical gate triggered

✖ EUAI-TECH-001 Logic pattern requires transparency markers (Art. 52).

✖ EUAI-DOC-001 Undocumented AI signal pattern detected in source.

The Engine identifies specific code-level markers and synthesizes the technical evidence required for regulatory documentation.

How Sentinel works

Sentinel converts regulatory requirements into deterministic checks that run inside your development workflow and produce verifiable compliance artifacts.

AI System

manifest.json

sentinel scan

policy engine

evidence artifacts

CI/CD gate

Protect your CI/CD pipeline

Block non-compliant AI systems before they reach production.

github-action.yml

name: Sentinel AI Compliance

on: [pull_request]

jobs:
  sentinel:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: MOXO08/sentinel-scan-action@v1
        with:
          manifest: manifest.json

Sentinel runs inside your CI pipeline and blocks pull requests that introduce regulatory violations or missing compliance artifacts.

Sentinel Compliance Check

EUAI-TECH-001 Transparency requirement missing

EUAI-DOC-001 Missing risk management documentation

Status: ❌ Compliance check failed

Surfaced directly in GitHub pull requests and Actions workflows.

Trusted by developers building compliant AI

Sentinel integrates directly into modern development workflows and CI pipelines.

Total scans executed

Global audit volume

Unique repositories

Verified Al Systems

Unique environments

Active CLI Installations

Built for engineering teams

Sentinel is an engineering-first tool designed to automate regulatory compliance within the development lifecycle.

CLI-first

Run deterministic compliance scans locally using a simple CLI. No vendor lock-in, no mandatory cloud sync for scanning.

CI/CD integration

Block non-compliant AI systems before they reach production. Automated status checks and PR comments for instant feedback.

Audit-ready evidence

Generate structured artifacts required for regulatory audits. Deterministic proof of compliance for every deployment.

The engineering layer for AI regulatory mapping

Sentinel provides the technical infrastructure to map software signals to regulatory patterns.

Scout: Discovery

Autonomous AI Signal Discovery
Automated Manifest Synthesis
AST-level Pattern Recognition
Continuous Inventory Crawling

Engine: Verification

Deterministic Rule Execution
WASM Sandbox Memory Isolation
Zero-Egress Technical Scans
SARIF Verification Reports

Hub: Governance

Immutable Audit Ledger (D1)
Cryptographic Evidence Fingerprints
Annex IV Document Synthesis
Sovereign Data Residency

Start building compliant AI systems today

Run Sentinel locally or integrate it directly into your development pipeline.

npx @radu_api/sentinel-scan

Run your first audit View on GitHub Read the documentation

Compliance Mapping

Built for AI compliance requirements

Sentinel helps engineering teams map AI system requirements to concrete compliance workflows, documentation artifacts, and audit evidence.

Art. 10

Data Governance

Support for documenting data governance, testing evidence, and dataset controls.

Art. 11

Technical Documentation

Generate structured technical artifacts and audit-ready compliance files.

Art. 12

Traceability

Maintain audit history, evidence records, and compliance traceability.

Art. 13

Transparency

Support documentation for instructions for use, disclosures, and oversight processes.

sentinel-evidence/

1 sentinel-evidence/

2 ├── compliance-summary.md

3 ├── scan-metadata.json

4 ├── scan-report.json

5 └── scan-report.sarif

Evidence output

Technical documentation artifacts generated

Technical Evidence for AI Regulatory Mapping