12 stories

Neo

Oh My Rogue Agent

Yesterday, Hugging Face came out saying they'd detected an AI autonomous-agent-powered cyberattack and that they had to use open-source models to actually investigate and remediate it. Later we heard from OpenAI that their agent was responsible; it happened during an ExploitGym eval, and the agent just drifted off the goal. It escaped the sandbox, reached OpenAI Research Environment, got access to internet, and hacked Hugging Face production environment trying to find the solution for the benchm

Continuous PR Security Review

The security findings that end up in incident post-mortems rarely looked dangerous in the PR that introduced them. Not because anyone was careless but because there's nothing in the change that looks wrong. The code does exactly what it says but the problem is in how the app behaves once it's running. A new endpoint ships without a permission check but every other route in the file handles permissions correctly, so nothing about it stands out. Or a response comes back carrying more of a user's

How Neo's Agent Architecture Evolved: From One Agent → Plan, Execute & Verify

Our first engineering post covered prompt caching, the infrastructure change that made long-running agentic tasks economically viable. That post assumed a multi-step, multi-agent system already existed. It did not exist on day one. When we started building Neo, the product was a single agent with a sandbox and a large toolset. Today, a typical task runs through optional planning, an Execution agent that delegates to parallel specialized subagents, and a verification loop that can re-run w

Red-Teaming Cloud Infrastructure with Neo

Most AI security tooling shipped over the last year focuses on one of two workflows, code review at PR time or zero-day research in open-source software. Models in PR pipelines now flag insecure patterns at every commit and autonomous research runs have produced more zero-days across open-source projects than the patch teams behind them can realistically triage. We've been running Neo on both of those workflows at ProjectDiscovery for a while now, surfacing zero-days in production software and t

Benchmarking Neo's Black-Box DAST Capabilities

Since the launch of Neo, we've been steadily expanding what it can do. Neo has found 33+ real CVEs across open-source projects, performed well on white-box security testing where source code is available, and generally proven itself as a capable security engineer when it has context to work with. What we hadn't shared yet is how Neo does when it's operating purely as a black-box DAST agent no source code, no architecture context, just a URL. The prompt Neo gets is a minimal prompt with no guida

Neo v. DIY: The gap between a single finding and a mature security program

In our latest webinar, our Founding Solutions Engineer, Davis Franklin, addressed the massive gap between finding a vulnerability with an LLM and running a mature security program. That gap is what Neo is built to close. With the release of Opus 4.6 and the announcement of Mythos, the question we hear constantly has gotten louder: Can I just build this with Claude Code? The short answer is yes. You can spin up a working PoC in about half an hour, find a real vulnerability, and feel genuinely co

How We Cut LLM Costs by 59% With Prompt Caching

At ProjectDiscovery, we've been building Neo, an autonomous security testing platform that runs multi-agent, multi-step workflows, routinely executing 20-40+ LLM steps per task. Vulnerability assessments, code reviews, and security audits at scale, enabling continuous testing across the entire development lifecycle. When we launched, our LLM costs were staggering. A single complex task with Opus 4.5 could consume 60 million tokens. Then we implemented prompt caching. Here's what changed:

Beyond the Model: Neo Hunts, Exploits, and Proves 22 Zero-Days.

LLMs are a genuine leap forward for vulnerability discovery. Anthropic reported 500+ zero-days from Opus 4.6 and OpenAI's Codex Security discovered 14 CVEs across projects like OpenSSH and GnuTLS. If you've experimented with LLMs for security testing, you've probably been impressed too. The practical reality for a security team deploying AI is messier than the headlines or early POC results suggest. Noise compounds fast. Anthropic brought in external security researchers to help validate the vo

Inside the benchmark: app architectures, walkthroughs of findings, and what each scanner actually caught

This is Part 2 of our vibe coding security benchmark study. In Part 1, we compared how LLM-based security tools like ProjectDiscovery's Neo and Claude Code performed against traditional SAST and DAST scanners on AI-generated code. We found that LLM-based tools like Neo and Claude Code detected many high-value findings that traditional scanners missed. Between Neo and Claude Code, Neo produced more true positives and fewer false positives because it could validate hypotheses against a running app

How Neo found an SSRF vulnerability in Faraday, and why it matters for every team that ships code

Executive Summary Neo found a Server-Side Request Forgery (SSRF) vulnerability in Faraday, a widely used HTTP client library in the Ruby ecosystem. This is Neo’s first credited CVE discovery. Neo is ProjectDiscovery’s AI security copilot for tasks like code review and vulnerability discovery. For this finding, Neo reviewed a widely used open source dependency and, without human guidance, surfaced a subtle URL-handling edge case, validated it in runtime, and produced a clear write-up that maint

AI code review has come a long way, but it can’t catch everything

AI code review can reason about intent, but real incidents often stem from business logic flaws that only show up in runtime. Our benchmark reveals where code-only review falls short.

Introducing Neo, an AI security engineer for complex security tasks

Neo is a cloud-based AI security engineer that works alongside your team and takes on real security tasks like a true co-engineer. As it operates, it continuously learns your systems and processes, improving over time just like an engineer ramping up on your team. Neo is built as a framework, not as a black box. It combines the reasoning of large language models with purpose-built execution tools, isolated sandboxes, a memory layer that learns your code, architecture and naming, and deep integr