> For the complete documentation index, see [llms.txt](https://whitepaper.gopluslabs.io/goplus-network/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://whitepaper.gopluslabs.io/goplus-network/overview.md).

# Overview

### Mission Statement

GoPlus is building the security layer for the AI era across two directions, AI for Security and Security for AI, with $GPS as the settlement unit for the security services that layer provides.

### What is GoPlus?

GoPlus is the open security layer that protects AI agents, Web3 transactions, and other high-risk digital actions before they execute.

AI agents are becoming active participants in digital economies. They browse the web, call tools, access files, manage credentials, write code, install packages, deploy software, and trigger transactions on behalf of users and organizations. As these agents scale, they will need security for runtime behavior, on-chain transactions, tool calls, credential access, code execution, package usage, plugin installation, and broader supply-chain dependencies.

Web3 remains one of the clearest examples of why execution-time security matters. Malicious signatures, phishing sites, scam contracts, rug pulls, wallet drainers, risky approvals, and irreversible transactions continue to threaten users, wallets, applications, and chains. The same execution risks will become more severe when AI agents can operate wallets, route trades, deploy contracts, and interact with dApps at machine speed.

GoPlus secures these environments through AI-powered risk intelligence, runtime protection, policy enforcement, verified security data, and open security services. Its mission is to make security a native layer of every autonomous action, whether that action is performed by a human user, an AI agent, a wallet, an application, or blockchain infrastructure.

Since 2022, GoPlus has grown from a Web3 user security infrastructure provider into a broad security network. Its security APIs process tens of millions of daily requests, cover more than 50 blockchain networks, protect millions of wallets, identify malicious assets and addresses at scale, and are integrated by wallets, dApps, exchanges, infrastructure providers, RPC services, rollups, and developer platforms. This ecosystem gives GoPlus a strong distribution foundation across supported networks, wallets, exchanges, DeFi protocols, launchpads, trading tools, data platforms, RPC providers, security partners, and infrastructure providers.

### Two Directions: AI for Security and Security for AI

GoPlus' strategy is organized around two directions that it intends to cover in full.

**AI for Security** uses AI to make security faster, deeper, and more accessible. GoPlus applies large language models, graph-based program analysis, dynamic fuzzing, financial semantic modeling, and continuously updated threat intelligence to the problems that have historically depended on scarce human expertise: auditing smart contracts, scanning software for supply-chain risk, detecting phishing and malicious addresses, simulating transactions, and classifying token risk. GoPlus DeepScan, GoPlus SafuSkill scanning, GoPlus Intelligence, and the Security Data Layer's computational verification are AI for Security. This direction matters because attackers are already using AI to lower the cost of vulnerability discovery and to scale fraud; defenders need AI to keep pace. GoPlus intends to release its own security-specialized language model, trained on the network's accumulated security data and red-team attack traces. The model serves both directions: it performs audit, scanning, and detection as AI for Security, and it drives red-teaming and red-versus-blue adversarial exercises against agent environments as Security for AI, with offensive findings training defensive capability. It will power GoPlus products and be consumed as a security service in its own right.

**Security for AI** secures AI systems themselves as they become autonomous actors. AI agents access models, tools, data, credentials, software, and financial infrastructure, and each of those touchpoints is a place where an agent can be manipulated, compromised, or made to cause harm. Axora Router, Axora AgentGuard, Axora Arena, SafuSkill's trusted distribution, and the AI Agent Security API are Security for AI: they govern the information flowing into agents, the actions agents take, the software agents install, and the on-chain interactions agents initiate.

The two directions reinforce each other. AI for Security produces the intelligence that Security for AI enforces at runtime, and Security for AI generates the machine-originated security consumption that funds the network producing that intelligence. Both directions deliver security as a service, and $GPS is the settlement unit across both.

| Direction           | What it means                                                                    | GoPlus products                                                                                                            |
| ------------------- | -------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------- |
| **AI for Security** | AI applied to security analysis                                                  | GoPlus DeepScan, GoPlus Intelligence, SafuSkill scanning, Security Data Layer verification, GoPlus Web3 Security detection |
| **Security for AI** | Security applied to AI systems and agents                                        | Axora Router, Axora AgentGuard, Axora Arena, SafuSkill trusted distribution, AI Agent Security API                         |
| **Across both**     | Security-specialized model for detection, red-teaming, and adversarial exercises | GoPlus Security Language Model (planned)                                                                                   |
| **Across both**     | Settlement, staking, incentives, governance                                      | $GPS                                                                                                                       |

GoPlus will continue to support its Web3 security business while extending the same security-layer philosophy to the AI era. Axora protects AI information flows and agent runtime behavior through Axora Router, Axora AgentGuard, and Axora Arena. SafuSkill secures the software supply chain that agents depend on. DeepScan provides AI-powered smart-contract auditing and continuous monitoring. GoPlus Intelligence and the AI Agent Security API power risk analysis for humans and machines alike. The GoPlus Security Network provides open security data, distributed security compute, and verifiable security signals. Security RPC, on-chain firewall, wallet integrations, and infrastructure integrations protect Web3 transaction execution. SafeToken Protocol and Locker standardize secure asset issuance and liquidity protection. $GPS connects all of them as the settlement and coordination unit for security-service consumption.

### The New Security Problem

Security risks are increasingly appearing at the moment of execution.

Traditional security systems were designed for applications that followed deterministic workflows. They monitored endpoints, scanned code, checked access control, and detected suspicious activity after systems were already built. AI agents and Web3 interactions change this model. Both are action-driven environments where software can make high-impact decisions, compose multi-step workflows, and trigger irreversible operations.

AI agents introduce a new runtime attack surface:

{% hint style="warning" %}
**Tool-call risk:** Agents can invoke shell commands, MCP tools, browser actions, cloud APIs, and deployment systems. A single unsafe tool call can leak secrets, modify production resources, or execute malicious code.
{% endhint %}

{% hint style="warning" %}
**Prompt injection and context manipulation:** Agents can ingest untrusted web pages, files, repositories, messages, and tool outputs. Malicious instructions hidden in those inputs can redirect agent behavior or override user intent.
{% endhint %}

{% hint style="warning" %}
**Credential and data exposure:** Agents often operate near sensitive assets such as `.env` files, private keys, API tokens, databases, customer data, and internal documents. Without runtime controls, sensitive data can be accessed, copied, or exfiltrated before a human reviewer notices.
{% endhint %}

{% hint style="warning" %}
**Agent supply-chain risk:** Agents rely on skills, plugins, MCP servers, packages, templates, repositories, browser content, and third-party tools. A compromised dependency or malicious integration can become an execution path into the agent runtime.
{% endhint %}

{% hint style="warning" %}
**Agent transaction risk:** AI agents will increasingly prepare, sign, route, simulate, and submit transactions on behalf of users or organizations. Without transaction-aware security, autonomous agents can trigger irreversible Web3 losses at machine speed.
{% endhint %}

Web3 user security faces a parallel execution problem:

{% hint style="warning" %}
**Irreversible transaction risk:** Once a malicious transaction or signature is submitted, users often cannot stop the resulting asset loss.
{% endhint %}

{% hint style="warning" %}
**High user decision burden:** Wallet warnings and raw transaction details still require users to understand complex contract behavior, approval logic, signature payloads, and token risks.
{% endhint %}

{% hint style="warning" %}
**Fragmented security coverage:** Users, applications, wallets, chains, and AI agents need consistent security policies across networks, tokens, dApps, RPCs, and transaction flows.
{% endhint %}

### Security as a Machine-Consumed Service

The shift to autonomous systems changes not only where security is needed, but how much of it is consumed.

Traditional security spend is seat-based and relatively fixed: an organization licenses software annually and a human operates it. Autonomous systems introduce a different pattern. An agent may require a security decision every time it accesses a model, invokes a tool, installs a skill, reads untrusted content, handles a credential, interacts with a counterparty, or executes a transaction. Security becomes a service that is consumed continuously, at machine speed, with usage scaling alongside agent activity rather than with the number of human users.

This is the addressable market GoPlus is expanding into. GoPlus is extending security services from human-initiated Web3 activity to autonomous software and AI agents, and positioning $GPS as the native settlement unit for that consumption. The economic logic is direct: more agents create more actions, more actions require more security checks, more security checks mean more security-service consumption, and more consumption is settled in $GPS. See [$GPS: The Security Service Settlement Layer](/goplus-network/usdgps/gps-security-service-economy.md) for the full model.

### GoPlus Security Layer

GoPlus addresses these risks by placing a security layer before execution.

Instead of relying only on alerts after an action has already happened, GoPlus evaluates risky actions in real time, applies policy, routes sensitive operations through approval when needed, blocks dangerous behavior, and records an auditable security timeline. This creates a common security model for both AI agents and Web3:

* **Detect:** Analyze prompts, URLs, commands, packages, skills, signatures, transactions, tokens, addresses, approvals, and dApps for risk.
* **Decide:** Evaluate each action against risk models, user intent, custom policies, and contextual security rules.
* **Enforce:** Allow, block, warn, or escalate high-risk actions before they execute.
* **Verify:** Use decentralized data contribution, distributed security compute, and validation mechanisms where transparency and trust are required.
* **Audit:** Produce clear records of actions, decisions, approvals, and security outcomes.
* **Settle:** Meter security-service consumption and settle it in $GPS, so that the operators and contributors who produce security are paid from the consumption they serve.

This security layer is open and modular. Developers can integrate GoPlus Intelligence through APIs and SDKs, and autonomous systems can consume it through the AI Agent Security API. AI agent users and teams can protect runtime behavior with Axora, supply-chain exposure with SafuSkill, and transaction execution with GoPlus Web3 Security. Wallets, dApps, RPCs, chains, and rollups can integrate GoPlus security capabilities through infrastructure integrations. Security developers, operators, and data contributors can participate through the GoPlus Security Network.

### Execution Security Architecture

GoPlus organizes execution security for autonomous systems into five layers. The first four are product capability layers, each covering a different point on the path from the software an agent depends on to the actions it ultimately executes. Axora forms the AI Agent Security product line; SafuSkill, DeepScan, and GoPlus Intelligence belong to the GoPlus Web3 Security product line. The fifth layer is the economic layer that runs across all of them.

| Layer                                        | Product                                       | Covers                                                                                                                                            |
| -------------------------------------------- | --------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------- |
| **01 — Software Supply Chain**               | GoPlus SafuSkill                              | Discovery, scanning, verification, and trusted distribution of the skills, scripts, and code agents rely on                                       |
| **02 — Agent Information Flows and Runtime** | Axora (Router, AgentGuard, Arena)             | AI security gateway and routing, runtime guard for agent actions, supply-chain scanning and daily patrol, and red-teaming of agent environments   |
| **03 — Smart Contracts and Applications**    | GoPlus DeepScan                               | Pre-deployment self-check, AI contract audit, and continuous security monitoring for the contracts agents and users interact with                 |
| **04 — On-chain Risk Intelligence**          | GoPlus Intelligence and AI Agent Security API | Token security, malicious address detection, transaction simulation, phishing detection, signature and approval analysis, for humans and machines |
| **05 — Security Service Settlement**         | $GPS                                          | Native settlement and coordination unit for security services consumed across the GoPlus ecosystem                                                |

### Core Product Lines

**Axora: AI Agent Security**

Axora is GoPlus' enterprise AI security product line, built around three control points on one control plane. **Axora Router** manages access: an enterprise AI security gateway that consolidates external AI access into one controlled entry point and enforces identity, group policy, sensitive-data controls, routing, and audit on every request and response. **Axora AgentGuard** protects action: an end-to-end agent security platform that discovers agents and their blast radius, validates real attack paths, and allows, warns, requires approval, or blocks agent actions at runtime from a deny-first control plane. **Axora Arena** validates the boundary: an adversarial security validation platform that runs authorized, controlled attack scenarios against AI systems and agents, produces reproducible findings, and feeds them back into defensive policy. Axora security decisions are metered security services that GoPlus intends to settle in $GPS.

**GoPlus Web3 Security**

GoPlus Web3 Security is the product line that protects the full lifecycle of on-chain activity and the software and contracts that lead to it. It includes GoPlus SafuSkill, GoPlus DeepScan, GoPlus Intelligence, transaction protection, and SafeToken Protocol, described below, as well as the GoPlus App, browser extension, GoPredict, Security RPC, wallet integrations, and on-chain firewall infrastructure that help users, wallets, dApps, chains, and RPC providers detect and prevent malicious on-chain actions before they execute. End users pay security gas in $GPS for transaction protection.

**GoPlus SafuSkill: AI Software Supply-Chain Security**

GoPlus SafuSkill is the trusted distribution layer for AI agent software. It aggregates AI skills, scans them for malicious commands, credential theft, and other supply-chain risks, and surfaces verified software to developers and agents through the GoPlus marketplace and partner distribution channels. Discovery, scanning, verification, and ongoing monitoring are each security services that GoPlus intends to settle in $GPS.

**GoPlus DeepScan: AI Smart-Contract Audit and Continuous Monitoring**

GoPlus DeepScan is the AI-powered smart-contract security product. It combines Graph-IR static analysis, LLM-powered semantic audit, dynamic fuzzing, financial semantic modeling, and accumulated security rule patterns to detect vulnerabilities, rug pull risks, access-control issues, honeypot behavior, and scam patterns. DeepScan covers the full contract lifecycle through Token Security Self-Check, AI Contract Audit, and Continuous Security Monitoring. Audits and monitoring subscriptions are security services settled in $GPS.

**GoPlus Intelligence and the AI Agent Security API**

GoPlus Intelligence is the risk analysis engine behind the security layer. It provides real-time security intelligence, transaction simulation, malicious address detection, token risk analysis, phishing site detection, approval and signature analysis, dApp risk information, and agent-oriented security detection. The AI Agent Security API exposes this intelligence to autonomous systems through a usage-based, machine-to-machine access model built on x402. GoPlus intends for $GPS to become the core settlement unit for this machine-originated consumption.

**GoPlus SafeToken Protocol and Locker**

SafeToken Protocol provides standardized security for token issuance and liquidity management through the SafeToken Factory and GoPlus Locker. It helps projects launch tokens with safer contract templates, risk controls, liquidity locking, and vesting, reducing malicious asset creation and improving trust across the token ecosystem. Service fees settle in $GPS, and $GPS is usable within partner token-management platforms such as Custos.

**GoPlus Security Network**

The GoPlus Security Network turns security into an open, verifiable, and participatory infrastructure layer. The Security Data Layer collects and verifies risk data from contributors, the Security Compute Layer distributes security analysis across a network of operators, and GoPlus Network Governance coordinates the ecosystem. Contributors and operators stake $GPS to participate and are rewarded in $GPS from the security-service consumption they serve.

### Why GoPlus?

**Security before execution:** GoPlus focuses on the moment where risk becomes damage: before an AI agent runs a command, before a tool call reaches an external system, before an agent or wallet submits a transaction, and before a user signs a malicious payload.

**Both directions of AI and security:** GoPlus is one of the few security networks positioned across AI for Security, using AI to audit, scan, detect, and simulate, and Security for AI, governing what agents see, do, install, and execute. The two directions share one intelligence layer and one settlement unit.

**Full-path coverage for autonomous systems:** SafuSkill, Axora, DeepScan, and GoPlus Intelligence cover the software agents depend on, the information and actions they process, the contracts they interact with, and the on-chain risk around their transactions.

**Proven Web3 security scale:** GoPlus has already built one of the most widely used Web3 user security infrastructures, with large-scale API usage, multi-chain coverage, wallet protection, malicious asset detection, and integrations across the ecosystem.

**Ecosystem distribution:** GoPlus works across supported networks, wallets, exchanges, DeFi protocols, launchpads, trading tools, data platforms, RPC providers, agent platforms, security partners, and infrastructure providers, giving the security layer broad distribution.

**Unified risk intelligence:** GoPlus combines AI-powered analysis, security data, transaction simulation, phishing detection, malicious address intelligence, token risk classification, and runtime risk detection into a common intelligence layer accessible to humans and machines.

**A token tied to service consumption:** $GPS utility is designed to scale with the security services the network actually delivers. As more agents take more actions that require more security checks, more consumption settles in $GPS and funds the operators and contributors that produce security.

**Open and verifiable network:** Through decentralized data contribution, distributed security compute, service coordination, and governance, GoPlus makes security services more transparent, scalable, and aligned with the communities they protect.

### Conclusion

The AI era requires a new security layer. As agents and applications gain the ability to execute complex actions on behalf of users, security must shift from passive detection to real-time protection before execution, and from seat-based software to a service consumed continuously by machines.

GoPlus is building that layer in both directions. As AI for Security, GoPlus applies AI to auditing, scanning, detection, and simulation through DeepScan, SafuSkill, and GoPlus Intelligence. As Security for AI, GoPlus governs agent information flows, actions, software, and on-chain interactions through Axora, SafuSkill, and the AI Agent Security API. Starting from its proven Web3 security foundation, GoPlus provides the intelligence, runtime controls, transaction security, supply-chain protection, policy enforcement, and verifiable services needed to secure high-risk digital actions across AI and Web3. $GPS is the settlement unit that ties the consumption of those services, in both directions, to the network that produces them.