01 Mission
Scanners find what is already known. We look for what isn’t.
ResilientX Research Labs is the offensive research team of ResilientX Security. We find exploitable vulnerabilities before attackers do, prove them end to end, disclose them responsibly, and are researching the AI that will do the same at machine speed.
01 Why a lab
Signature-based security is bounded by the CVE feed
Scanners, signatures and compliance checklists can only confirm what someone has already found. The vulnerabilities that end up in breach reports are, on the day they are first used, in none of them.
Our work starts below that line: taint analysis from attacker-controlled input to dangerous sinks, binary diffs of vendor patches, snapshot fuzzing of firmware and exploit development against real mitigations, until a hypothesis becomes a deterministic proof of concept.
Every result follows the same pipeline: vendor first, then a CVE record and a technical advisory, then a detection or exposure check in the ResilientX Security platform. Next, an AI model that runs that pipeline at machine speed, with researchers in the loop.
02 Principles
How we work
Impact over volume
One deterministic exploit is worth more than a hundred unverified scanner findings. We prove the primitive before we rate the bug.
Vendor first
Every finding goes privately to the vendor. We publish once users can protect themselves, and we do not sell vulnerabilities or exploits.
Show the work
Advisories carry the root cause, CWE, CVSS v4.0 vector, detection guidance and the full timeline, so defenders can verify instead of trust.
Machines for scale, people for judgement
Automation and AI take the repetitive work at machine speed. Researchers own targeting, exploitation and every disclosure decision.
03 Research
What we work on
- 01
Zero-day research
Original vulnerability research in edge appliances, enterprise software, parsers and cloud control planes: the pre-auth attack surface where a single bug is a breach.
Explore - 02
Exploit development
Proving impact instead of guessing it: from a crash to a primitive, from a primitive to code execution, against real mitigations on production configurations.
Explore - 03
Disclosure & CVEs
Vendor first, always: private reports to the vendor’s PSIRT, coordinated embargoes, CVE records through a CNA, and advisories with root cause and detection guidance once a fix ships.
Explore - 04
Threat research
How real attackers operate: n-day weaponisation timelines, exposed infrastructure at internet scale and new exploitation techniques, turned into detections and exposure checks.
Explore - 05In research
Offensive AI
An AI model built for offensive security, in active research: it reads decompiled code, writes fuzzing harnesses, triages crashes to root cause and plans exploitation, and every hypothesis it makes is verified by execution.
Explore
04 Method
From hypothesis to advisory
- 01
Map
Enumerate the attack surface: pre-auth endpoints, IPC and RPC interfaces, file and protocol parsers, and the privilege each trust boundary runs with.
- 02
Hunt
Taint-guided code review, CodeQL and Semgrep variant analysis, binary patch diffing, structure-aware and snapshot fuzzing under sanitizers.
- 03
Prove
Minimise the crash, establish the primitive, and build a deterministic proof of concept against a release build with default mitigations on.
- 04
Disclose
Private report to the vendor’s PSIRT with root cause, PoC and CVSS v4.0 vector; coordinate the fix, the CVE record and the embargo.
- 05
Publish
Full advisory once patched: root cause, affected and fixed versions, detection guidance and the complete disclosure timeline.
05 Engage
Request an
engagement
Free for open-source products
Commission the Labs team for white-box or black-box research on a product, device or service you build or depend on: zero-day hunting, exploitability assessment or threat research. Tell us the target and we will scope the work with you.
Found a vulnerability in our products? Report it privately to security@resilientx.com · security.txt