zerocrit
Penetration testing services

Penetration testing that drives your criticals to zero.

Zero crit is pentester shorthand for a report with no critical findings left open. Zerocrit is built to get you there, with autonomous scanning that never stops looking and a penetration tester who digs in by hand. It is led by a bug bounty hunter with 6+ years in the field and multiple CVEs credited. You get one quote, one fixed price and one report, priced to cost less than running a bug bounty program.

  • Manual + autonomous
  • CVSS-scored findings
  • Retest included
  • Wildcard scopes
Criticals open0
Click to break it
  • 6+ yearsof penetration testing and bug bounty hunting
  • CVEscredited for vulnerabilities found in real products
  • OWASP Top 10flaws found and reported in production systems
  • Fixed priceagreed in writing before any testing starts
Try it

Get this sample report to zero crit.

These are sample findings from a made-up engagement. Fix each one and watch the retest clear it.

2
critical findings open
    Coverage

    Eight attack surfaces under one scope.

    Attackers do not stop at the edge of a web app, so a penetration test should not either. Add the ones you run to your scope and we test each one against a written methodology.

    Web applications

    Authentication, access control, injection, business logic and session handling across every role in your app.

    OWASP WSTG · OWASP Top 10

    APIs

    REST, GraphQL and gRPC. Broken object and function level authorization, mass assignment and rate limits.

    OWASP API Security Top 10

    Mobile apps

    iOS and Android binaries, local storage, certificate pinning and the backend calls they make.

    OWASP MASVS · MASTG

    Cloud

    AWS, Azure and Google Cloud. IAM privilege paths, exposed storage, metadata abuse and misconfigured services.

    CIS Benchmarks · IAM review

    External network

    Everything you expose to the internet, including the forgotten subdomains and services nobody remembers launching.

    Asset discovery · PTES

    Internal network

    What happens after one laptop is compromised. Active Directory attack paths, lateral movement and privilege escalation.

    Active Directory · Kerberos

    AI and LLM features

    Prompt injection, data leakage, unsafe tool calls and broken guardrails in chatbots, agents and RAG pipelines.

    OWASP Top 10 for LLM Apps

    Source code

    Assisted code review that follows risky data flows the black-box test pointed at, plus secrets and dependency checks.

    SAST · SCA · secrets
    • Multiple scopesPut a web app, its API and your cloud account in one engagement and one quote.
    • Wildcard scopesGive us *.example.com and we map the subdomains and test what we find.
    • All in writingYour quote lists every scope before testing starts.
    Method

    Machines find the breadth. People find the depth.

    Scanners are fast and tireless but they cannot understand what your product is supposed to do. Testers can, but their hours are limited. We use each for the work it does best and bill you once.

    Attack pathSample · click to replay
    High 8.1ZC-014Sample finding

    Any customer can read any other customer's invoices

    CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N · GET /api/v2/invoices/{id}

    Found by: manual testing. The scanner mapped the endpoint. A tester noticed the invoice ID was never checked against the signed-in account.

    Autonomous scanning

    Always running

    Our scanning pipeline maps your attack surface and checks it continuously, so new exposures are caught between manual tests.

    • Subdomain, port and service discovery
    • Known CVEs and misconfigurations
    • Leaked secrets and exposed admin panels
    • Re-checks after every change to your perimeter
    Manual penetration testing

    Thinking like an attacker

    A tester works through your application by hand and chains small weaknesses into real impact, the way an intruder would.

    • Business logic and authorization flaws
    • Multi-step exploit chains
    • Proof of concept for every finding
    • Scanner results reproduced before they reach you
    1. Scope and authorize

      We agree targets, dates and rules of engagement. Nothing is tested until you sign the authorization.

    2. Autonomous sweep

      Scanning maps every in-scope asset and flags the quick wins.

    3. Manual testing

      Testers go deep on the areas that matter and try to break what the scanners could not.

    4. Report and walkthrough

      You receive the report and a call to go through each finding with your engineers.

    5. Retest

      Once you have fixed the issues, we test them again and update the report.

    Founder

    Run by someone who hunts bugs for a living.

    I'm Xiety. For more than six years I have broken into systems with permission, as a penetration tester and a bug bounty hunter.

    In that time I have reported vulnerabilities to many companies before attackers could use them. That work has earned multiple CVE credits and uncovered flaws from across the OWASP Top 10 in live production systems.

    I started this company because I have worked on both sides. Bounty programs pay hunters per bug, so the crowd goes where the rewards are. Your first serious test should look at everything in scope and cost exactly what was quoted.

    whoamiFounder
    alias
    Xiety
    role
    Founder, lead penetration tester
    experience
    6+ years, pentesting and bug bounty
    cves
    Multiple credited
    findings
    Across the OWASP Top 10, in production
    Zerocrit vs. bug bounty platforms

    A bounty program bills you for what it finds. We bill you once.

    Bug bounty platforms are a good fit for some teams. They are an unpredictable way to buy your first thorough test. Here is how the two models differ.

    Bug bounty platformZerocrit pentest
    What you payAn annual platform fee, plus a bounty pool, plus fees on payouts, plus optional triage.One fixed price, agreed before testing starts.
    Budget certaintySpend rises with the number of valid reports.Known on day one. A bad result costs the same as a clean one.
    CoverageResearchers choose what to look at. Unpopular assets can go untested.Every in-scope asset is tested against a written methodology.
    NoiseDuplicate and low-quality reports need filtering.Each finding is reproduced by a tester before you see it.
    Audit evidenceA stream of individual reports. A formal pentest is usually bought separately.A signed report and attestation letter for your auditor and customers.
    Best forMature security teams that want many researchers looking all year.Teams that need a thorough test, a clear report and a fixed budget.

    The two are not rivals for every team. Many mature companies run both. If you are choosing which to buy first, a penetration test gives you full coverage and a known cost.

    Pricing

    Prices you can read before you talk to anyone.

    Prices are in US dollars. Your final quote depends on scope and is fixed in writing before work begins.

    Sweep

    Autonomous scanning

    $490 per month

    Continuous scanning of your external attack surface, with a person checking results.

    • Unlimited external assets on one domain
    • Alerts for new exposures
    • Findings reviewed by a tester
    • Monthly summary report
    Start scanning
    PentestStart here

    Manual + autonomous

    $4,900 from, per engagement

    A full penetration test of one web application or API, with scanning throughout.

    • Manual testing by a named tester
    • Autonomous sweep of the whole scope
    • Report with CVSS-scored findings
    • Walkthrough call with your engineers
    • One retest and attestation letter
    • More scopes or a wildcard on the same quote
    Get a fixed quote
    Continuous

    Testing all year

    $2,400 per month

    Quarterly manual tests with always-on scanning between them.

    • Four manual tests a year
    • Everything in Sweep
    • Testing of major releases on request
    • Unlimited retests
    • Shared channel with your testers
    Talk to us
    The report

    Written for engineers, readable by auditors.

    The report is the product. It tells your developers exactly what to fix and gives your auditor and your customers the evidence they ask for.

    • Part 1
      Executive summaryOverall risk, what was tested and what needs attention first, in two pages.
    • Part 2
      Scope and methodologyTargets, dates, accounts used and the standards followed.
    • Part 3
      FindingsEach with a CVSS 3.1 score and vector, steps to reproduce, evidence and a specific fix.
    • Part 4
      Retest resultsThe status of every finding after your fixes.
    • Letter
      Attestation letterA one-page summary you can share with customers. Useful as evidence for SOC 2, ISO 27001 and PCI DSS requirement 11.4.
    Questions

    Penetration testing, answered plainly.

    How much does a penetration test cost?
    A Zerocrit penetration test of one web application or API starts at $4,900. Larger scopes such as several applications, mobile apps, cloud accounts or internal networks cost more. You receive a fixed quote in writing before any work begins, and the price does not change with the number of findings.
    Can one engagement cover multiple scopes or a wildcard domain?
    Yes. One engagement can hold several scopes, such as a web application, its API and a cloud account. Wildcard scopes like *.example.com are welcome: we map the subdomains and test what we find. Your quote lists every scope in writing before work begins.
    How is Zerocrit different from a bug bounty platform like HackerOne or YesWeHack?
    Bug bounty platforms connect you with a crowd of independent researchers and you pay a platform fee plus a reward for each valid report. Zerocrit is a penetration testing company. Our own testers examine every in-scope asset against a methodology for one fixed price, and you receive a single report. A bounty program suits mature teams that want ongoing crowd attention. A penetration test suits teams that need complete coverage and a predictable cost.
    What is the difference between manual and autonomous penetration testing?
    Autonomous testing uses software to discover assets and check them for known vulnerabilities and misconfigurations, continuously and at scale. Manual testing is done by a person who understands how your product should behave and looks for logic flaws, broken access control and exploit chains that software misses. Zerocrit uses both on every engagement.
    Will the report satisfy SOC 2, ISO 27001 or PCI DSS?
    The report and attestation letter are written to serve as penetration testing evidence for SOC 2, ISO 27001 and PCI DSS requirement 11.4. Your auditor makes the final decision, so tell us which framework you are working toward and we will scope the test to match it.
    Is it safe to test our production systems?
    Yes, with care. We agree rules of engagement before starting, avoid destructive actions and denial-of-service techniques, and can test a staging environment if you prefer. If we find a critical issue during testing we tell you the same day and do not wait for the report.
    How long does a penetration test take?
    Most single-application tests take one to two weeks of testing, followed by the report within five business days. Larger scopes take longer. We confirm dates when we send your quote.
    Do you retest after we fix the findings?
    Yes. Every penetration test includes one retest of the reported findings within 60 days, and the report is updated to show what has been fixed. Continuous plans include unlimited retests.
    Get a quote

    Tell us what you run. We reply with a fixed price.

    1. Pick what you want tested
    2. Add your targets. Several at once and wildcards are fine
    scope.txt0 targets
      Try one:
      3. Send your request
      or
      4. Or email it yourself
      scope@zerocrit.com

      Only list systems you own or are authorized to test. Add any deadline or compliance framework. We reply within one business day with a fixed price.