proofrange@labs:~$ ./hire --real
ProofRange
  • Platform
  • How it works
  • Pricing
  • FAQ
  • LoginTry It Now
  • Platform
  • How it works
  • Pricing
  • FAQ
  • LoginTry It Now
ProofRange

ProofRange is a technical assessment service by CloudBreach, a cloud security company specialising in offensive security and hands-on training.

Quick Links

  • Home
  • About Us
  • Pricing
  • Contact Us

Other Pages

  • Terms & Conditions
  • Privacy Policy

© 2026 ProofRange, powered by CloudBreach

proofrange@labs:~$ ./assess --real

Cyber skill, proven on real systems.

Put candidates and teams into real environments: cloud labs, terminals, remote desktops, and live web targets, plus question sets for the theory. AI scores the work and an integrity check runs in the background. Built for hiring, and for growing the team you already have.

Start free →Contact us
Free trialNo credit cardHiring and upskilling
proofrange · privilege escalation · linuxlive
candidate@target:~$ sudo -l
(root) NOPASSWD: /usr/bin/find
candidate@target:~$ sudo find . -exec /bin/sh \; -quit
# id
uid=0(root) gid=0(root) groups=0(root)
root@target:~# cat /root/flag.txt
flag{priv_esc_via_sudo_find}
proofrange> submit flag
flag accepted. challenge solved.

Integrity

92/100
Low risk
0 pastes, 1 tab switch

AI evaluation

Strong Read the sudo rights, spotted the find GTFOBin, and got root cleanly. Efficient, no brute force.
Typed, not pasted
1 tab switch, 28 seconds
Built to hire and grow
PentestersSOC analystsCloud securityAI securityDFIRRed teamSecurity engineers
The problem

Claims are easy. The work is the test.

A resume tells you what someone says they can do. A quiz tells you what they remember. Neither shows whether they can sit down at a real system and get the job done. ProofRange measures both. Ask questions for the knowledge, and put candidates in a live environment for everything a question cannot show.

proofrange@labs:~$ ./real-environments

Give them a real system, not a screenshot.

Each hands-on challenge runs in its own isolated environment that the candidate actually operates, right in the browser with nothing to install. When the job is done, they capture a flag to prove it.

☁

Cloud labs

Full cloud environments to build, attack, or defend.

>_

Terminal access

A real shell on the box. Recon, exploit, escalate.

▢

Remote desktop

RDP and VNC in the browser. No client, no setup.

⟠

Live web targets

Real, vulnerable web apps to probe and break.

terminal · web exploitation
$ nmap -sV 10.10.0.5
22/tcp open ssh
80/tcp open http Apache 2.4.49
$ curl --path-as-is 10.10.0.5/cgi-bin/../../etc/passwd
root:x:0:0:root:/root:/bin/bash
$ searchsploit Apache 2.4.49
Path Traversal and RCE (CVE-2021-41773)
vuln confirmed. building the exploit
proofrange@review:~$ ai-evaluate

AI reads every attempt, so your team does not have to.

Open answers and hands-on sessions are summarized and scored automatically. Every candidate is measured against the same bar, and a review that used to take an evening takes a few minutes. You stay in charge: the AI points out what matters, and you make the call.

Strongtier verdict
Minutesnot evenings
One barfor everyone
AI review · Priya N. · web exploitation
Overall assessmentStrong

Chained a path traversal into remote code execution on the Apache target, then moved to the database with reused credentials. Clear method, little wasted effort, each step written up. Two challenges solved in full, one partial.

2 / 3challenges solved
41 mintime used
4 / 5open answers
proofrange@integrity:~$ score --explain

Integrity is built in. No webcam needed.

Every assessment gets a score from 0 to 100, built from real interaction data: how much was pasted, how often the tab lost focus, idle time followed by sudden bursts, and typing rhythm. AI flags sit on top. A flag asks a person to take a closer look. It never rejects anyone on its own.

integrity report · attempt 4821
63/100
Medium, review suggested
Typing rhythm matches genuine hands-on work
3 large pastes during the coding task
5 tab switches, 2 longer than 30 seconds
No idle then burst pattern

Two modes, set per assessment

▷ Focus mode: monitoring and scoring in the background, nothing for the candidate to fight🔒 Strict mode: fullscreen enforced, copy and paste blocked, every exit logged

Use Focus for a calm candidate experience, or switch any assessment to Strict when more is on the line. Either way, you can see the exact reasons behind every score.

Build

Build the assessment that fits the role.

Pick from a library of hands-on challenges and question banks, or write your own. Run it as questions only, hands-on only, or a mix of both. Set the time limit, the difficulty, and the integrity mode for each one, then invite candidates one at a time or in bulk.

Questions onlyHybridHands-on only
AddItemTypeDiff.Points
✓Linux Privilege Escalationhands-onMed100
✓Network Recon BasicsquestionEasy40
✓Path Traversal to RCEweb targetHard150
+SOC: Triage an Alertremote desktopMed120
Decide

Rank people on what they actually did.

Scores, time, and integrity in one sortable table. Sort by integrity to send your review time straight to the attempts that need a second look.

#CandidateScoreTimeAI EvalIntegrity
1PNPriya N.8641mStrong92 Low
2DMDiego M.8152mPromising88 Low
3YHYara H.7438mBorderline63 Review
4TVTomas V.6960mLimited90 Low
Two jobs, one platform

Hire from outside. Grow from within.

⚲

Hiring

Test candidates on real skill before or after the interview. Choose the format that fits the role: questions only, hands-on only, or both. Walk into the hiring panel with evidence instead of opinions.

↗

Upskilling

Benchmark your own team, find the gaps, and track progress over time. Turn a rough sense of where the team needs work into a plan you can measure.

How it works

From job spec to decision in five steps.

01

Build

Pick challenges and questions, set the mode and integrity.

02

Invite

Add candidates one by one or in bulk.

03

They work

Candidates work in a real, isolated environment.

04

AI scores

The work is summarized, scored, and integrity checked.

05

You decide

Rank, compare, and choose with evidence.

Pricing

Plans that grow with your hiring.

Start free, then move to a paid plan when you are ready. Switch between monthly and yearly anytime.

Credits are used when a candidate begins an assessment, so you only pay for real attempts.
Loading plans...
Need more credits? Boost packs add credits that stay valid for 12 months, on any paid plan.
FAQ

Questions, answered.

How are credits counted?
A credit is used when a candidate or employee starts an assessment, not when you assign it. Invite as many people as you like and only pay for the ones who begin. Top-up credits stay valid for 12 months and are used only after your plan credits run out.
Is this proctoring?
No proctoring. There is no webcam and no recording. Integrity is a score from 0 to 100 built from interaction data, with AI flags on top. It points reviewers to attempts worth a closer look, and never rejects anyone on its own.
Do candidates need to install anything?
No. Cloud labs, terminals, remote desktops, and web targets all run in the browser. A candidate opens the link and starts working.
Can I run questions instead of hands-on?
Yes. Every assessment can be questions only, hands-on only, or a mix of both. You choose the format for each one.
Can I use it for my own team?
Yes. The same challenges and scoring work for internal upskilling. Benchmark your team, find gaps, and track growth over time.

Run your first assessment, free.

Build an assessment, invite a candidate, and watch real evaluation in action. No credit card to start.

Start free →Contact us