Pharos Taps CertiK Skill Scanner for AI Agent Hackathon With 150,000 PROS Prize Pool

Pharos Network has selected CertiK’s newly launched Skill Scanner as a core evaluation tool for its global AI Agent Hackathon, a six-week developer initiative designed to accelerate the creation of secure, production-ready AI agents.

The competition forms part of Pharos’ broader “Create Like a PRO · Agent Carnival” campaign, which features a total prize pool of 150,000 PROS tokens, worth approximately $92,500 at current prices.

Unlike traditional hackathons that focus primarily on building end-user applications, the Pharos event is structured around a “Skills first, Agents second” framework. Participants are first tasked with creating reusable AI Skills — modular capabilities that can later be assembled into autonomous agents capable of handling on-chain transactions, payments, collaboration, content creation, and other blockchain-based activities.

The strongest projects will compete for a dedicated 50,000 PROS reward pool while gaining exposure to future opportunities across the Pharos ecosystem.

Loading tweet...

View Tweet

Moving Beyond Demo-Only AI Agents

As interest in autonomous AI agents continues to grow, many projects remain largely experimental, with limited real-world utility beyond demonstrations. Pharos believes the next generation of agents must be able to operate independently, manage assets, interact with on-chain protocols, and participate in digital economies over extended periods.

Building that future requires more than just powerful models. It also requires standardized skills, secure execution environments, and robust verification frameworks capable of evaluating how agents behave in production settings.

That is where CertiK’s Skill Scanner enters the picture.

Security Becomes Part of the Judging Criteria

Launched in May, Skill Scanner is CertiK’s security assessment platform designed specifically for AI Skill marketplaces, developers, and enterprise environments.

The tool allows users to submit Skills through GitHub repositories, URLs, or ZIP packages and scans for risks including:

• Malicious code execution

• Data leakage

• Unauthorized network access

• Shell execution vulnerabilities

• File system abuse

• Runtime security risks

Unlike conventional static code analysis tools, Skill Scanner is designed to evaluate risks that may emerge while software is actively running, an increasingly important consideration as AI agents begin handling financial transactions and autonomous decision-making processes.

Its inclusion in the Pharos judging process marks a notable shift in how AI hackathons evaluate projects.

Historically, judging criteria have centered on innovation, usability, and technical execution. By integrating security directly into the scoring framework, Pharos is signaling that secure design should be considered a foundational requirement rather than something addressed after launch.

Loading tweet...

View Tweet

Security Standards Emerging Across AI Ecosystems

CertiK says Skill Scanner is already seeing adoption across multiple AI-focused ecosystems, including AI Skill marketplace Pieverse and AI assetization platform FinChip.ai.

As AI agents move closer to real-world deployment across finance, payments, and blockchain infrastructure, security tooling is increasingly becoming a critical layer of the emerging agent economy.

The Pharos hackathon may offer an early glimpse into how future AI ecosystems evaluate developers — rewarding not only innovation, but also the ability to build agents that can operate safely in production environments.