Now available for Shopify, WooCommerce & BigCommerce

MCP Gateway • HITL • Tool Governance

Your AI agents make thousands of decisions per hour. Do you know what they're doing?

The Universal Governance Layer for AI Agents

Enforce policies, block dangerous tools, and require human approval across AI frameworks and e-commerce platforms

LangChain•CrewAI•AutoGen•OpenAI•MCP•CLI•Shopify•WooCommerce•BigCommerce•Any Framework•Custom Agents•LangChain•CrewAI•AutoGen•OpenAI•MCP•CLI•Shopify•WooCommerce•BigCommerce•Any Framework•Custom Agents

Get Started Free View on GitHub

Try Demo - No Signup Required

50+PII types|

13countries|

136integrations|

49MCP tools|

8published SDKs

tork-pii-demo

InputEdit to try your own text

Tork Output

Click 'Redact PII' to see the magic...

See Full Demo

MCP-Native Security

Govern Every Agent, Every Tool Call

Tork sits between your AI agents and the outside world. Every MCP tool call, every API request, every database query passes through governance before it executes.

🔒

MCP Gateway

Purpose-built for Model Context Protocol. Scan MCP servers, govern tool calls, and enforce policies across Claude, Cursor, and Windsurf agents.

👤

Human-in-the-Loop

Require manual approval for high-risk actions. Your agents propose, your team decides. Slack, email, and dashboard integrations.

🛑

Kill Switch

Instantly halt any agent with a single API call. Block all tool calls, notify your team, and generate an audit trail — in milliseconds.

📋

Policy Engine

Define JSON rules for what agents can do. Block dangerous tools, restrict data access, and enforce least-privilege across your fleet.

🔍

PII Detection

50+ PII types across 13 countries. Zero configuration — automatic region detection via IP and content signals. Custom PII patterns available on Enterprise.

📜

Compliance Receipts

Every governance decision generates a cryptographic, hash-chained audit entry. Hand these to auditors for HIPAA, GDPR, SOC 2, or PCI-DSS.

Try the Interactive Demo

Works with 136+ integrations

LangChainCrewAIAutoGenOpenAIMCPAnthropicLlamaIndexFastAPIHugging FaceMistralSemantic KernelAWS Bedrock

How It Works

Three Steps to Governed AI

Define Policies

Write human-readable YAML policies.

policies:
  - name: pii-protection
    action: REDACT
    patterns: [email, phone]

Wrap Your Calls

Add Tork to your AI pipeline.

from tork_governance import Tork
tork = Tork()
result = tork.govern(output)

Monitor & Audit

Get dashboards and receipts.

# Every evaluation creates
# a signed receipt:
{
  "decision": "ALLOW",
  "hmac": "a3f2c1..."
}

Get Started

Install in Seconds

8 published SDKs (Java and Swift in progress), a CLI, and an MCP server

Python

PyPI: tork-governance

$ pip install tork-governance

v0.23.0PyPIPython 3.10+

JavaScript

npm: tork-governance

$ npm install tork-governance

v0.10.0npmNode 18+

v0.1.0

$ go get github.com/torkjacobs/tork-go-sdk

GitHubGo 1.21+

Rust

v0.2.0

$ cargo add tork-governance

crates.ioRust 1.70+

Ruby

v0.2.0

$ gem install tork-governance

RubyGemsRuby 3.0+

PHP

v0.1.0

$ composer require tork/governance

PackagistPHP 8.1+

.NET

v0.1.0

$ dotnet add package TorkGovernance

NuGet.NET 8.0+

Java

—

Not yet published

Maven CentralJava 11+

Swift

—

Not yet published

SPMSwift 5.9+

Elixir

v0.1.0

{:tork_governance, "~> 0.1.0"}

Hex.pmElixir 1.14+

Tork CLI

npm: @torknetwork/cli

$ npx @torknetwork/cli

Governed in 4 commandslogin · init · test · doctorCLI Quickstart →

MCP Server

npm: @torknetwork/mcp-server

$ npx -y @torknetwork/mcp-server

49 Governance ToolsHITL

Works with Claude Desktop, Cursor, and any MCP-compatible client

View documentation on GitHub →

Zero Code Required

Configure Your Way

Three ways to manage AI governance. Pick what works for you.

No Code

Admin Console

Point and click. Embed our React dashboard directly in your app.

<TorkAdminConsole apiKey="..." />

Developers

SDK & API

Full programmatic control. Python, JavaScript, or REST API.

tork = Tork()

AI Teams

AI-Assisted

Ask Claude or Cursor to configure governance for you.

"Register a tool that blocks --force flag"

All three methods write to the same database. Changes apply everywhere instantly.

Features

Core Tools for AI Governance

Production-ready compliance without the complexity

MCP Gateway

49 governance tools via Model Context Protocol. Drop-in integration with Claude, Cursor, and MCP clients.

Learn more →

Tool Call Governance

Block dangerous flags, enforce allowlists. Rate limiting and target environment restrictions.

Learn more →

HITL Systems

Human approval with fatigue detection. Velocity alerts catch slicing attacks.

Learn more →

Policy Enforcement

Define rules in YAML, enforce everywhere. Block, warn, or allow based on content.

Learn more →

PII Redaction

Auto-detect and redact 50+ PII types across 13 countries. Zero configuration — region detected from request IP and content signals. Enterprise: custom pattern builder with test harness.

Learn more →

Compliance Receipts

HMAC-signed, tamper-evident audit trails. Designed to help demonstrate compliance.

Learn more →

Listen Mode

Deploy in observation mode first. Tork observes and records your traffic patterns automatically.

Learn more →

Human-in-Loop

Route sensitive decisions to humans. Configurable escalation paths.

Learn more →

Framework Agnostic

Works with LangChain, CrewAI, AutoGen, or raw API calls.

Learn more →

Tork significantly reduces AI safety risks through multiple detection layers. No AI safety solution guarantees 100% protection. We recommend using Tork as part of a defense-in-depth strategy. See our Terms of Service for details.

Differentiators

Why Tork?

What sets us apart from the competition

Memory Integrity

Unique

Designed to monitor agent memory drift. Helps identify when agents may deviate from safety instructions or exhibit unexpected behavioral changes.

Framework Agnostic

Universal

Works with LangChain, CrewAI, AutoGen, OpenAI, Anthropic, and any custom framework. One governance layer for all.

MCP Native

49 Tools

49 governance tools via Model Context Protocol. Drop-in integration with Claude Desktop, Cursor, and any MCP client.

Kill Switch CLI

Instant

Stop any agent instantly with tork agent kill. Freeze for investigation, resume when safe. Full audit trail.

OpenClaw Security

OpenClaw is powerful.
Tork makes it safe.

Potentially malicious skills exist in the MCP ecosystem. Tork Guardian is designed to scan, flag, and help govern skill interactions — reducing risk for your agents.

🛡️ PII Redaction🔒 Shell Blocking⚡ Port Security

$ npm install @torknetwork/guardian

Available on ClawHub — clawhub install tork-guardian

Learn More

Trusted by developers building AI

3,784+

Tests Passing

A+

Security Headers Rating

~20ms

Added overhead at the edge — cloud governance typically completes in under a second

136

Framework Adapters

50+

PII Types Detected

MIT

Open Source License

Protocol Layer

Govern every agent protocol

Tork intercepts and governs at the protocol layer — not just content. Every message inspected, every decision receipted.

A2A

Agent-to-Agent

Governance for agent-to-agent communication. Every A2A message is inspected, redacted if needed, and receipted before delivery.

ACP

Agent Communication Protocol

Full ACP message governance with PII detection, policy enforcement, and cryptographic audit trails.

AG-UI

Agent UI Protocol

Govern the bridge between agents and user interfaces. Ensure no sensitive data surfaces in UI streams.

Consent

Consent Protocol

Enforce consent policies across your agent stack. Every interaction logged, every consent decision receipted.

POST /api/v1/govern/a2a

{
  "content": "Transfer $5,000 to account 4111-1111-1111-1111",
  "agent_id": "planner-01",
  "agent_role": "planner"
}

Protocol Docs

ARCH Intelligence

Runtime intelligence built in

Tork learns your agent fleet's behavior, enforces budgets, caches results, and alerts you when something changes.

ARCH-05

Swarm Budget

Per-agent and per-org daily/monthly call limits enforced in Redis. Budget exceeded? The call is blocked before it reaches the LLM.

ARCH-06

Tool Result Cache

SHA-256 content-addressed caching for governed tool calls. Identical inputs return cached results instantly — zero redundant LLM spend.

ARCH-07

Behavioral Baseline

Learns normal call patterns per org — volume, agents, PII types, active hours. Flags anomalies with a 0-100 score in real time.

ARCH-08

Daily Diff Email

Automated daily report comparing today vs yesterday. New agents, volume spikes, PII changes — delivered to your inbox at 7am UTC.

Tork Chat

AI-Powered Customer Experience — Governed by Default

Deploy a fully governed AI assistant on your website in minutes. Multi-agent intelligence, compliance built in.

Multi-Agent Intelligence

7 specialized LangGraph agents route every conversation to the right expertise — fleet, policy, pricing, booking, escalation — automatically.

Built-In Governance

Every message passes through Tork’s L3/L6 compliance pipeline. PII detection, audit receipts, and escalation controls — no extra setup.

5-Minute Deployment

Upload your documents, customize your widget, copy one line of code. Your AI assistant is live on your website in minutes, not months.

💬

Value Car Hire

Welcome to Value Car Hire! 👋 How can I help you today?

What SUVs do you have available in Cape Town?

Great question! We have the Haval Jolion Pro (R449/day) and GWM P-Series 4X4 (R549/day) available in Cape Town...

Type a message...

🔒

PII Detection

50+ PII types auto-detected across 13 countries — no region configuration required

📋

Compliance Receipts

Cryptographic proof for every AI interaction

🤖

7 Specialized Agents

Not a generic chatbot — purpose-built AI for your industry

⚡

Sub-second Responses

Streaming SSE for instant, token-by-token delivery

Start Free See Demo

Trusted by Value Car Hire · RentWorks Africa

Protocol Layer

Protocol-Native Governance

First-class support for every major agent communication protocol

A2AGoogle

Live

Agent-to-Agent

Google’s agent communication protocol — governed with PII detection and cryptographic receipts.

/api/v1/govern/a2a

ACPCisco / Agentio

Live

Agent Communication Protocol

Cisco/Agentio standard for agent networks — policy enforcement and audit logging on every exchange.

/api/v1/govern/acp

AG-UICopilotKit

Live

Agent-User Interaction

CopilotKit’s frontend protocol — every user-facing AI interaction scanned before it reaches the screen.

/api/v1/govern/agui

ConsentTork

Live

Consent Protocol

Tork’s consent verification layer — ensure every AI action has proper authorization with tamper-proof receipts.

/api/v1/govern/consent

Every protocol message is scanned for PII, checked against your policies, and logged with an auditable governance receipt — before it reaches your agents.

📖 Free Book

The Agent Crisis

The definitive guide to AI governance. Free.

94 pages covering why multi-agent systems need governance, how to implement it, and what happens when you don't. Co-authored by the Tork founding team.

Download Free Read Online

ISBN 978-1-7644797-2-1Free · No signup required

Testimonials

What Governed AI Looks Like

Real governance scenarios powered by Tork.

“Deployed Torky in under a week. Our after-hours enquiries are now handled instantly.”

Value Car Hire

Car Rental · Cape Town, South Africa

“The governance layer gave us the confidence to deploy AI customer-facing. Full audit trail, zero compliance concerns.”

RentWorks Africa

Fleet Management · Johannesburg, South Africa

“A customer service AI agent was leaking email addresses and phone numbers in its responses. Tork’s PII detection layer caught and redacted the sensitive data before it reached end users.”

PII Detection in Action

Customer Service AI Use Case

These scenarios illustrate common governance use cases. Results vary based on configuration and deployment.

Customer Spotlight★★★★★

“We run AI agents across product creation, SEO, inventory and customer support. Before Tork, we had no visibility into what those agents were doing with sensitive data. Now every action is governed, auditable, and we sleep a lot better at night.”

Founder, Roar 4x4

JCorp Australia · Automotive E-Commerce

PII Incidents

<5min

Integration

100%

Audit-Ready

View all reviews →

Case Studies

Deployed Across Industries

From car rental to financial services to education — governed AI at every scale.

Vehicle Rental

Live

AI assistant handling fleet queries, pricing, booking, and insurance — 24/7 with governance.

70%+ enquiries automated · Sub-2s response · Lead capture built in

South Africa

Financial Services

Proposal Stage

Compliance-first AI governance for banking and insurance. PII detection, audit trails, POPIA + GDPR aligned.

136 adapters · Cryptographic receipts · Multi-framework support

Africa

Education

In Development

AI-powered learning across 2,600+ textbook titles. RAG retrieval for institutional knowledge bases.

Pearson partnership · ASEAN market · B2B + B2C

ASEAN

The Standard

The TORKING-X Standard

Beyond tooling — a proposed framework for measuring AI governance quality.

Transparency

Full visibility into agent decisions

Oversight Compliance

Regulatory alignment tracking

Risk Level

Real-time risk-level scoring

Kill Switch Response

Emergency stop capability

Integrity Score

Data and process validation

Network Trust

Multi-agent coordination safety

Governance Adherence

Policy compliance tracking

eXplainability

Decision reasoning clarity

Pricing

Start Free, Scale with Confidence

Simple pricing for teams of all sizes. All plans include core governance features.

MonthlyAnnualSave up to 34%

FREE

$0 USD

For testing and evaluation

5,000 API calls/month
Basic PII detection
1 team member
Community support
7-day log retention

Get Started

STARTER

$29 USD/mo

Billed $348 USD/year — save 26%

50,000 API calls/month
120 requests/minute
Advanced PII detection
3 team members
Email support
30-day retention

Enterprise Solutions

Custom governance infrastructure for large organizations. Dedicated support, SLAs, and white-glove onboarding.

Starting at $2,000 USD/mo

See Enterprise Plans

Unlimited API calls & team members
Dedicated support & SLA
SSO, SAML & custom domains
Blockchain-anchored audit trails
Multi-region deployment options
On-premise deployment (contact us)

All prices in USD. Local currency conversion available at checkout via Stripe.

Complimentary with your Tork account

The Agent Crisis

How to Build a Safer, Smarter, and More Prosperous World with Autonomous AI

By Yusuf Jacobs · 94 pages · 21 chapters

A practical guide to AI governance. From PII redaction to compliance receipts, learn how to build governance systems that actually work.

Learn More Download Free

Comparison

Why Tork?

The only governance platform built for the multi-agent era.

CapabilityOthersTork

Agent coverage1–5 frameworks136 adapters, 8 published SDKs, a CLI + e-commerce platforms

Compliance proofLogs onlyCryptographic HMAC-signed receipts

PII detectionBasic regexGravity-weighted multi-layer detection

Protocol supportNoneA2A, ACP, AG-UI, Consent — all live

Integration timeDays to weeks3 lines of code

PricingEnterprise onlyFree tier available

Ready to Govern Your AI?

Join developers building trustworthy AI systems with Tork.

Start Free Trial Star on GitHub

Get AI governance insights weekly — no spam, unsubscribe anytime.

The Universal Governance Layer for AI Agents

Govern Every Agent, Every Tool Call

MCP Gateway

Human-in-the-Loop

Kill Switch

Policy Engine

PII Detection

Compliance Receipts

Three Steps to Governed AI

Define Policies

Wrap Your Calls

Monitor & Audit

Install in Seconds

Configure Your Way

Admin Console

SDK & API

AI-Assisted

Core Tools for AI Governance

MCP Gateway

Tool Call Governance

HITL Systems

Policy Enforcement

PII Redaction

Compliance Receipts

Listen Mode

Human-in-Loop

Framework Agnostic

Why Tork?

Memory Integrity

Framework Agnostic

MCP Native

Kill Switch CLI

OpenClaw is powerful.Tork makes it safe.

Govern every agent protocol

Agent-to-Agent

Agent Communication Protocol

Agent UI Protocol

Consent Protocol

Runtime intelligence built in

Swarm Budget

Tool Result Cache

Behavioral Baseline

Daily Diff Email

AI-Powered Customer Experience — Governed by Default

Multi-Agent Intelligence

Built-In Governance

5-Minute Deployment

PII Detection

Compliance Receipts

7 Specialized Agents

Sub-second Responses

Protocol-Native Governance

Agent-to-Agent

Agent Communication Protocol

Agent-User Interaction

Consent Protocol

The Agent Crisis

What Governed AI Looks Like

Deployed Across Industries

Vehicle Rental

Financial Services

Education

The TORKING-X Standard

Start Free, Scale with Confidence

Enterprise Solutions

The Agent Crisis

Why Tork?

Ready to Govern Your AI?

OpenClaw is powerful.
Tork makes it safe.