> stop threats in real time
Scan your website, code and servers, then let HackEx keep watch. When a real threat shows up, it explains the problem in plain language and ends it: containment on your approval or fully automatic, and code fixes delivered as pull requests.
// capabilities
From first scan to stopped attack
Deep Code Scanning
Not pattern matching · over two thousand AST rules across PHP, Python, JS/TS, Go, Java, Ruby, C#, Rust and more, plus cross-file taint tracking that follows user input through your own functions into the dangerous call. Verified secrets, vulnerable dependencies with reachability, and Terraform, Kubernetes, Docker and CI-pipeline checks.
Live Watch
Checks your site every minute. If it goes down, AI figures out why and restarts it for you (automatically with Auto Mode, or on your approval), with the whole timeline logged.
Threat Stop + Auto Mode
One click blocks an attack. Or pick your autonomy: smart mode acts on strong evidence, full bypass runs every whitelisted action instantly. Everything auto-reverses and stays audit-logged.
Code Fixes & Deploy
HackEx writes the patch itself and opens a pull request on GitHub or a merge request on GitLab. You review and merge, or grant deploy permission and it merges and pulls the fix onto your server.
Compliance & Evidence
Findings map to SOC 2, PCI DSS and ISO 27001 with a readiness score and SLA clocks. Auditors get one-click CSV exports · findings, incidents, scans and a hash-chained audit trail that proves nobody tampered with the record.
Enterprise Controls
SAML single sign-on through Okta or Entra, per-project access scoping for team members, a CI/CD gate that fails builds on blocking findings, ticketing webhooks for Jira or ServiceNow, and a CycloneDX SBOM per scan.
// meet_hex
HEX, your security co-pilot
Not a chatbot bolted on. HEX sits inside your workspace, reads your real scans, incidents and watch states, and answers in plain English by text or live voice. It can also act: run scans, arm watch, propose or execute containment.
- > talk or type: realtime voice calls, or plain chat
- > knows your stack: answers from your data, not generic docs
- > hands, gated by you: manual approval, smart mode or full bypass
fig. 01 · hex_core // signal integrity 55%
// pipeline
Connected to protected in minutes
Paste a URL, connect GitHub or GitLab, or run one command on your server. Takes minutes.
We check everything once: open doors, weak locks, missing protections, scored out of 100.
Then we keep watching. Anything new (a weakness, an attack, an outage) pings your team instantly.
HackEx opens a code-fix PR for your review, and contains live attacks on your approval or fully on its own in Auto Mode.
// zero-trust, by design
$ config --get secrets_masked
Passwords and keys we spot are hidden instantly: never stored, never sent anywhere.
$ config --get no_code_at_rest
Your code is scanned then deleted. The evidence stays on your servers, not ours.
$ config --get audit_everything
Every action is logged forever. Even our own admins only see metadata, never your content.
> sudo protect --everything
Free scan now. Continuous watch, AI write-ups and Threat Stop when you create a workspace.