Script-Kiddie — AI Security Workbench

Everyone's a script kiddie now.
We make it safe.

Open-source tools and labs for modern AI security practitioners. Govern, test, and deploy AI — without the risk.

guardrails.ts

scan.py

policy.yaml

1// GuardRails — Secure AI Paste Engine

2import { scan, govern } from '@sk/guardrails'

3import { OWASPCheck } from '@sk/owasp'

5async function secureAI(input: string) {

6 const result = await scan(input)

7 // Strip secrets, PII, injections

8 result.filter(['secrets', 'pii', 'injection'])

10 const policy = await govern(result)

11 if (!policy.compliant) {

12 throw new ComplianceError(policy.violations)

13 }

15 return await orchestrate(result)

16}

// Platform

Built for enterprise security

Every line of code passes through a hardened security and governance pipeline.

🔍

Paste & Scan Engine

Real-time scanning strips secrets, PII, tokens, and injection vectors before anything is processed.

ZERO-TRUST INPUT

🏛️

AI Governance Layer

Policy-as-code governance. Define what models, data, and outputs are acceptable — per team.

POLICY-AS-CODE

🌐

Multi-Model Orchestrator

Routes across GPT-4o, Claude, Gemini, DeepSeek and more. Picks the best output automatically.

BEST-OUTPUT ROUTING

// Products

Three ways to engage

Free · Open Source

⚔️ SK Framework

Offensive AI security training

✓ Modular attack framework
✓ Prompt injection labs
✓ OWASP ML Top 10 exercises

Clone on GitHub →

Enterprise

🏢 GuardRails

Secure AI coding & orchestration

✓ Secure paste & code scanning
✓ AI Governance & policy-as-code
✓ Compliance (GDPR, HIPAA, SOC2)
✓ SSO, RBAC, Audit Logs

Request Demo →

Services

📊 SK Services

Assessments & Training

✓ Red-team assessments
✓ Governance framework design
✓ Team training & certification

Everyone's a script kiddie now.We make it safe.