Find, Fix, Repeat.Secure your No flaws, no bill.

VibeDefend wires your coding agent to CybeDefend in one command. Business rules mined from the way you code.

Install in 5 secondsNode 18.17+

npx -y @cybedefend/vibedefend@latest install

Auto-detects

Claude Code
Cursor
OpenAI Codex
Windsurf
VS Code Copilot

The coding agents we wire into

Run the installer once; it auto-detects every agent on your laptop and wires the ones you pick.

Why flaws ship

Your agent writes the line. Nobody reads it.

Thousands of lines a day. No review catches the flaw.

Shipped with the flaw

Claude CodeAgent session, alone

Thinking ...

Update ( src/db/users.repo.ts )

+db . query('SELECT * FROM users WHERE id = ' + id)

Tests passing

Opening pull request #482

» auto modeVibeDefend: off

A rules file does not fix it.

The rule was in its file, exact. The agent broke it anyway.

M06 · LOY-01Loyalty pointsThe rule was in its file

The company ruleOne loyalty point per euro paid in money. None on the gift-card share.

ViolatedA rules file

What VibeDefend changes

Two minutes to install.

One command. It detects every agent on your machine.

$npx -y @cybedefend/vibedefend@latest install

Claude Code

Cursor

Windsurf

GitHub Copilot

OpenAI Codex

Your rules arrive at the line.

Mined from your code, plus OWASP, SOC 2, GDPR, ISO 27001.

Claude Code

VibeDefend

vibedefend . business-rules.inject

tenant scopingsoft-deleteaudit-on-write

vibedefend . security-rules.inject

OWASPSOC 2GDPRISO 27001

Thinking ...

Update ( src/db/users.repo.ts )

+db . query('SELECT * FROM users WHERE id = ' + id)

vibedefend . scan.diff

Code AnalysisIaCSecrets

new finding in diff → SQL Injection

» auto modeVibeDefend active

Scanned before the pull request.

One finding, with its fix. The agent rewrites the line.

src/db/users.repo.tsdiff scanned in the session

db.query('SELECT * FROM users WHERE id = ' + id)

db.query('SELECT * FROM users WHERE id = $1', [id])

SQL Injection, fixed at the line it edited

Rule LOY-01 applied exactly

Dangerous commands never run.

A schema drop, a destructive sudo. Stopped before they fire.

Agent session · guarded

$sudo rm -rf /etc/

BLOCKEDdestructive sudo outside the project· no-destructive-sudo

$DROP SCHEMA public CASCADE

BLOCKEDschema drop against a live database· no-destructive-sql

$requests-toolbelt-async

BLOCKEDpackage does not exist on the registry· package does not exist

Every finding, live in the session.

One dashboard for your team. The same list, inside the agent.

CybeDefend project view with live findings — Every finding, with its fixThe agent reads this list live

What you get

Four scenes, recorded.

Real Claude Code sessions with VibeDefend in the loop. Pick one.

Claude CodeVibeDefend

Add an endpoint to update user profile

Four bricks, one install.

What runs inside the agent once the command has run.

Rules

Business rules mined from your code, plus OWASP, SOC 2, GDPR and ISO 27001, injected with the prompt.

vibedefend · rules.inject

tenant scopingsoft-deleteaudit-on-writeno-raw-amount-math

OWASPSOC 2GDPRISO 27001

Secret guard

Raw secret reads are refused; the agent gets the managed reference instead.

vibedefend · action.guard

$cat .env | grep STRIPE⊘blocked raw secret read (.env) · logged to audit→reference process.env.STRIPE_WEBHOOK_SECRET

$constructEvent(body, sig, process.env.STRIPE_WEBHOOK_SECRET)✓npm test -- webhooks · 4 passed

Action guard

Destructive commands, schema drops, protected paths: stopped before they fire, logged to the audit trail.

vibedefend · action.guard

$rm -rf ./storage/uploads⊘protected path · outside allowed scope

$DELETE FROM sessions⊘no WHERE · 8,932 rows · logged to audit

Live findings

Every scanner, unified. The agent reads the full findings list live and fixes in the loop.

vibedefend · findings.list

SASTSQL injection · users.repo.ts:42 · reachable
SCAlodash 4.17.20 · CVE-2021-23337
secretsAWS access key · config/prod.yml
IaCS3 bucket public-read · main.tf
CI/CDpull_request_target · .github/workflows/deploy.yml

in the agent's context · live

Eight flaws, stopped before the first save.

43% of API vulnerabilities exploit business logic, not CVEs (Wallarm, 2026). No scanner catches them.

Claude Codebefore

1// GET /api/orders/:id2const order = await db.order.findUnique({3  where: { id: req.params.id },4});5return res.json(order);

VibeDefendafter

1// GET /api/orders/:id2const order = await db.order.findUnique({3  where: { id: req.params.id, tenantId: req.user.tenantId },4});5if (!order) return res.status(404).end();6return res.json(order);

Lookup by id with no tenant filter: any signed-in customer can read this order.

Caught at agent-time · Claude Code

Behind VibeDefend, a whole platform.

What you install in the agent runs on the full CybeDefend platform: dashboard, scanners, policies, reports and AutoFix, for the whole team.

VibeDefendin your agentThe CybeDefend platformDashboardScannersPoliciesReportsAutoFixSee the full platform

Built to be trusted with your code.

GDPRCompliant

Compliance is effective today. EU customers run on SecNumCloud-qualified infrastructure.

SOC 2 Type IIControls in placeAudit under way

Our controls meet the criteria and the audit is under way. The attestation ships under NDA the day it is issued.

Trust Center

Your code stays in your regionEU in Paris, US in Iowa

The guard decides locally. Analysis stays in your region, on our own models, and never trains one.

Partners, programs and investors

Built by experts

Jason LeeFormer CISO, Zoom and Splunk. Virtual CISO at F5.

Twenty years at the top of security, from thirteen years at Microsoft to SVP Security Operations at Salesforce. He joined the board in September 2026.

**Florentin Ledy**Co-founder, Ops & Tech at CybeDefend

Read the study and reproduce it

Free to start. The whole platform.

No card, no time limit.

€0no card

10 static scans
50 AI credits
Access to the whole platform
Your first scan unlocks the promo code

Prefer to see it on your repo first?

Twenty minutes with a founder, on your repo.

Book a demo Read the study and reproduce it36 pages, 90 transcripts, blind audits.

How it works,
in detail.

The documentation walks you through setup, configuration and every option.

Read the docs

Agent & AI Integration

VibeDefend, one command to wire every agent

The one-command install, the MCP server and hooks, linking a project, every command, tunables and troubleshooting.

Before you install.

Does my code leave my machine?

What you send for analysis does, and it stays in the region you picked at install: Paris for EU customers on SecNumCloud-qualified infrastructure, Iowa for US customers, on our own self-hosted models, never a third-party AI API. The guard decides on your machine, its telemetry carries metadata only (no prompt content, no raw code), findings never cross regions, and your code is never used to train a model.

Which agents does it work with?

The installer wires Claude Code, Cursor, Windsurf, GitHub Copilot and Codex. Any other agent that speaks MCP works in bring-your-own-config mode, including Gemini CLI, Cline, Continue.dev and Zed.

What does the free plan include, and what happens after?

50 AI credits and 10 static scans, with access to the whole platform, no card and no time limit. Run your first scan and we mail you a promo code. When you need more, the Developer plan starts at €19 or $21 a month.

Is this another scanner that floods the agent with alerts?

No. Legacy scanners run on commits and bury you in false positives. VibeDefend runs at agent-time on the diff being written, with reachability and framework awareness, and hands the agent one finding it can act on, with the fix.

Can I read the study before I sign up?

Yes, and you should. The 36-page paper, the 90 transcripts, the blind audits and the six cases where the layer did not deliver are in the public repository linked on this page.

Nobody told your agent. Now it knows.

One command. Two minutes. No card.

Install VibeDefend in 5 seconds.

One command wires every coding agent on your machine to CybeDefend: your business rules, your compliance frameworks, and guards that block destructive calls before they fire.