YFarmX logoYFarmX

Tools

The Toolkit

The Model Compare Bench leads: two AI models, measured against each other string by string. Then Trackers watch what is going wrong, Indexes catalogue what exists, Utilities do one job fast, and the Glossary explains the words. Every number below is live from the same data the tools run on.

283
AI models measured
$5.8bn
exploit losses logged
3050
terms decoded

The bench

The most-used tool on the site: two models, measured against each other.

Trackers

The Security Desk’s live logs: what is going wrong, as it breaks.

security desk92AI incidents loggedAI Risk RadarThe emerging risks of frontier AI, logged as they break: prompt injection, jailbreaks, hijacked agents, poisoned models, deepfake fraud and AI-found bugs, with severity and a source on each.security desk$5.8bntracked across 140 exploitsCrypto Exploit TrackerEvery major crypto exploit: what was hit, how much, and how, with on-chain references and sources on each incident.security desk2035global migration deadlineQuantum Threat TrackerThe honest state of the quantum threat: the Q-Day estimate, who has migrated to quantum-safe crypto, whether the top-100 coins are ready (none are yet), and the hardware milestones.security desk$4,000median Bugcrowd ceiling · 1,178 programmes trackedBounty EconomicsWhat public bug bounty programmes pay at their ceiling, how fast they answer, who gives safe harbour, and each programme's rule on AI-assisted reports, measured from the platforms' own data on seven platforms.security desk47vulnerabilities confirmed at a recordAI-Found Vulnerability RegisterEvery CVE a public record credits to an AI finder, with the tier of evidence beside the credit: the record's own credits array, the CNA's advisory, or a claim outside the record, with claimed identifiers and aggregate counts held apart.security desk30models scored on security workModel Security CapabilityHow each frontier model does at finding, reproducing and writing up vulnerabilities, with the harness, the task count and the safeguard state beside every score, and what each lab's model will answer.

Indexes

The structured reference databases behind each desk.

Utilities

Fast, single-purpose tools, live from the same data they run on.

Reference

The definition layer under every article on the site.