JuriSync Documentation

A development-time cross-jurisdictional compliance orchestration layer and Model Context Protocol (MCP) server for AI-assisted software engineering.

As developers build AI-enabled features using autonomous coding agents (Claude Desktop, Cursor, Antigravity), systems are deployed across international borders without automated visibility into regulatory divergence. A system built in the US may function fine under voluntary guidelines like the NIST AI RMF 1.0, but upon international release trigger mandatory high-risk conformity assessments under EU AI Act Annex III, or violate India's DPDP Act Section 8(7) data erasure rules.

Discovering these non-conformities during post-launch legal reviews forces costly architectural rewrites, schema overhauls, or regulatory fines. JuriSync shifts regulatory evaluation left into the earliest design phase before code leaves your editor.

Quickstart in 5 Minutes

JuriSync can be utilized in three primary modalities:

Modality 1: Interactive Web Console

Visit the live interactive console at https://jurisync-one.vercel.app/. Enter a natural language feature description (or select one of the 4 preset scenarios) and press Ctrl + Enter to generate an instant regulatory blueprint.

Modality 2: Zero-Install MCP Server (`npx`)

Connect JuriSync directly to Claude Desktop, Cursor, or any agentic IDE with zero cloning:

{
  "mcpServers": {
    "jurisync": {
      "command": "npx",
      "args": ["-y", "jurisync"]
    }
  }
}

Modality 3: Terminal Headless Stdio Call

Execute the compliance pipeline from any shell via stdio:

echo '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"analyze_compliance","arguments":{"feature_description":"AI system filtering resume CVs in EU and India"}}}' | npx -y jurisync

Core Engineering Principles

1. Deterministic Rule Matching

Every finding emitted by JuriSync corresponds to a deterministic rule object containing exact statutory citations (e.g. eu_ai_act_employment_high_risk, in_dpdp_retention_erasure_mandate). Rules are curated and mapped to statutory triggers (`employment`, `biometric`, `health`, `autonomous_action`, `personal_data`, `children`).

2. Binding vs. Voluntary Stratification

Regulatory confusion often stems from treating voluntary guidance as binding law, or vice-versa. JuriSync strictly classifies all output:

• [BINDING LEGISLATION]: Enacted statutes carrying legal enforcement mechanisms, statutory damages, or regulatory fines (e.g., EU AI Act, GDPR, India DPDP Act 2023, NYC Local Law 144, Illinois BIPA).
• [VOLUNTARY FRAMEWORK]: Industry risk management standards and federal guidance (e.g., NIST AI RMF 1.0, EEOC guidance).

3. Zero-Egress Privacy Architecture

JuriSync is engineered with zero remote telemetry. Web console analyses run client-side in the browser; MCP stdio queries run locally in Node.js. No developer prompts, proprietary feature descriptions, or candidate resumes are ever transmitted to third-party servers.

npm Package & CLI Specification

JuriSync is distributed on the official npm registry as jurisync (v1.1.0).

# Install locally in a project
npm install jurisync

# Or run instantly via npx without installing
npx -y jurisync

Model Context Protocol (MCP) Setup

The Model Context Protocol (MCP v1.0) is an open specification enabling AI coding assistants to invoke external tools over JSON-RPC 2.0 stdio pipes.

Claude Desktop Configuration

Locate your configuration file:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "jurisync": {
      "command": "npx",
      "args": ["-y", "jurisync"]
    }
  }
}

Cursor IDE Configuration

In Cursor, navigate to Cursor Settings → Features → MCP and click Add New MCP Server:

• Name: jurisync
• Type: command
• Command: npx -y jurisync

Node.js SDK Usage

Use JuriSync directly in your CI/CD linters, GitHub Actions, or backend compliance microservices:

import { orchestrateCompliance } from 'jurisync';

const feature = 'Automated candidate screening tool using CV parsing and facial video interview analysis in the EU and India';

// Run central compliance orchestration pipeline
const report = orchestrateCompliance(feature);

console.log('Signals Detected:', report.signals.tags.map(t => t.label));
console.log('EU Rules Triggered:', report.guidance.EU.rules.length);
console.log('India Rules Triggered:', report.guidance.IN.rules.length);
console.log('Cross-Border Tensions:', report.tensions.map(t => t.title));

// Export Markdown action checklist for GitHub PR
console.log(report.checklist);

5-Stage Orchestration Pipeline

When a query is received, JuriSync executes five deterministic stages in sequence (< 5ms total execution latency):

Risk Signal Extraction Radar (15 Signal Domains)

The signal engine detects 15 granular regulatory risk classes with phrase n-gram matching, compound clause negation handling (e.g., "does not process biometrics or facial scans"), and a 0–100% confidence rating:

Cross-Border Tension Engine (6 Modeled Divergences)

The Tension Engine surfaces architectural friction where complying with one country's statutory obligation creates compliance challenges under another's, complete with severity ratings and engineering explorations:

Curated Statutory Coverage Matrix (32 Rules)

Every rule in JuriSync is verified against official government publications. Filter by jurisdiction below:

Model Context Protocol (MCP) Tool Reference

JuriSync exposes three standard JSON-RPC 2.0 tools over stdio:

1. analyze_compliance

Runs the complete orchestration pipeline on a natural language feature description.

{
  "name": "analyze_compliance",
  "arguments": {
    "feature_description": "An AI recruitment system that parses employee resumes and scores candidates for interview rejection in the EU and India.",
    "jurisdictions": ["EU", "IN", "US"] // Optional, defaults to all
  }
}

2. get_statutory_rule

Retrieves the full statutory citation, legal classification, and developer action for a specific rule ID.

{
  "name": "get_statutory_rule",
  "arguments": {
    "rule_id": "eu_ai_act_employment_high_risk"
  }
}

3. list_supported_regulations

Lists all 15 tracked international statutory instruments and regulatory bodies.

Responsible AI & Governance Framework

Compliance automation requires strict architectural ethics:

• Decision Support vs. Replacement: JuriSync is designed for engineers and architects to flag risks during sprint planning. It empowers informed dialogue with legal counsel.
• Human-in-the-Loop Safeguards: Where AI makes automated evaluations affecting individuals, JuriSync enforces oversight mechanisms per GDPR Art. 22 and EU AI Act Art. 14.
• Data Minimization: The system processes only the natural language feature description provided. No system code, applicant PII, or internal credentials are required or ingested.

Engineering Roadmap (6 Phases)

The ongoing development roadmap includes: