- Alerts without a confidence model
- Context split across tools
- Reports that stop at “possible”
BUILT FOR PEOPLE WHO TEST REAL SYSTEMS
Find the holes.
Keep the proof.
HOLES turns scanner signals into controlled verification and evidence you can reproduce, explain and act on.
First release in progress. Private beta applications are open.
- 01DiscoverMap the real surface●
- 02VerifyTest within explicit limits●
- 03ProvePreserve reproducible evidence●
AFTER THE SCANNER FIRES
Cut the noise.
Follow the evidence.
HOLES takes the signal forward: verify it within explicit limits, preserve the trail and hand over a finding another engineer can reproduce.
- Evidence state always visible
- Scope, identity and test pressure attached
- A reproduction path ready to hand off
THE WORK STARTS AFTER DETECTION
Know what matters.
Know what to do next.
A severity score can start the conversation; evidence finishes it. HOLES keeps the operator close to the signal, the verification path and the decision.
Separate signal from finding.
Keep observed, suspected and confirmed states distinct, so attention follows the evidence rather than the loudest alert.
CONFIDENCE MODEL · EVIDENCE STATESCarry the proof into remediation.
Keep reproduction steps, technical context and the verification trail attached to the finding from discovery to fix.
REPRODUCIBLE STEPS · DECISION CONTEXTTune pressure to the target.
Set intensity, request ceilings and allowed actions to match target sensitivity. Move deeper only when scope and operator approval allow it.
ADAPTIVE INTENSITY · OPERATOR CONTROLPut AI on the reasoning path.
AI connects signals and proposes the next bounded check. Registered verifiers run it; reproducible evidence determines confidence.
AI REASONING · DETERMINISTIC VERIFICATIONTHE ASSESSMENT STAYS VISIBLE
See the reasoning.
Control the pressure.
Define the target, choose test intensity, watch each check and keep evidence tied to the decision that produced it.
Define the target
application.example
09:42:18surface 18 endpoints mapped
09:42:22policy target remains inside scope
09:42:29verify collecting deterministic evidence
EVIDENCE FIRST
Confidence has to be earned.
Observed, suspected or confirmed: HOLES keeps the distinction visible so the next decision starts from what the evidence actually supports.
Cross-user file access
A controlled A/B test confirmed that one authorized identity can read a synthetic resource belonging to another.
BUILT FOR AUTHORIZED DEPTH
Push where it matters.
Keep control where it counts.
Real targets. Explicit permission.
Work on applications and networks you own or are authorized to assess. Scope, identity and request ceilings travel with every check, so depth never means losing control.
AUTHORIZED VERIFICATION · RECORDED SCOPE · REPRODUCIBLE PROOFAI proposes.
HOLES verifies.
AI reads the trail, connects signals and proposes the next check. Registered verifiers execute it; higher-impact actions wait for operator approval.
AI REASONING · CONTROLLED TOOLS · HUMAN AUTHORITYPOWER UNDER OPERATOR CONTROL
Choose the depth.
Keep the proof.
Pressure that fits the target
Choose intensity and request ceilings for the system in front of you, then increase depth when the evidence justifies it.
AI-guided. Verifier-executed.
AI shapes the next move. Registered tools execute it. Deterministic evidence sets the confidence level.
Evidence ready to move
Traceable results and encrypted evidence packages stay useful through review, handoff and remediation.
TECHNICAL QUESTIONS
The technical picture.
What is HOLES.pro?
HOLES is an evidence-first workspace for mapping web and network attack surfaces, verifying signals and turning the result into a reproducible technical trail.
Where does HOLES fit in a pentest workflow?
It takes the workflow beyond discovery: define scope and test pressure, verify signals with registered checks, classify confidence and preserve the evidence trail.
How does HOLES handle confidence and false positives?
Observed, suspected and deterministically confirmed findings remain separate. Each state carries its scope, verification steps and supporting evidence.
What does AI do inside HOLES?
AI connects signals, reads the evidence trail and proposes the next bounded check. Registered verifiers execute the test and reproducible evidence determines confidence.
Where does the operator stay in control?
You set authorization, scope, identity, test pressure and request ceilings. Higher-impact actions remain behind explicit approval.
Who is HOLES for?
Pentesters, AppSec engineers, security consultants and technical teams who need findings they can reproduce, explain and move into remediation.
Can HOLES verify vulnerabilities on real systems?
Yes—on systems you own or are explicitly authorized to test. Test pressure, scope, identities and request limits stay attached to the assessment.
What should beta testers challenge?
Push the setup, assessment workflow, test-pressure controls, evidence quality, reproducibility and handoff. We want the friction, not polite approval.
When will it be available?
Private beta applications are open. We onboard small cohorts as builds become ready and move toward the first release.
GET EARLY ACCESS
Bring your workflow.
Push ours.
HOLES is moving toward its first release.
If you work in pentest, AppSec or security engineering, help us sharpen the checks, evidence and operator experience.