Skip to main content
Vulnerability skills provide Strix agents with advanced testing techniques for specific vulnerability classes. Each skill includes exploitation patterns, bypass techniques, validation methods, and false positive identification to help agents conduct thorough security testing.

​
Available Vulnerability Skills

SQL Injection

Union, blind, error-based, and ORM bypass techniques. Covers MySQL, PostgreSQL, MSSQL, and Oracle primitives with modern JSON/JSONB exploitation.

XSS

Reflected, stored, and DOM-based XSS testing with CSP bypass techniques. Framework-specific sinks for React, Vue, Angular, and Svelte.

IDOR

Object-level authorization failures (BOLA/IDOR) for cross-account data access. GraphQL, WebSocket, and multi-tenant testing patterns.

Authentication (JWT)

JWT and OIDC security testing covering token forgery, algorithm confusion, and claim manipulation across microservices.

Business Logic

Workflow bypass, state manipulation, and domain invariant violations in financial, e-commerce, and SaaS contexts.

CSRF

Cross-site request forgery testing including token bypass and SameSite exploitation patterns.

SSRF

Server-side request forgery covering internal service access, cloud metadata endpoints, and protocol smuggling.

RCE

Remote code execution through deserialization, template injection, and unsafe expression evaluation.

Path Traversal

Local file inclusion, remote file inclusion, and directory traversal with encoding bypass techniques.

XXE

XML external entity attacks for file disclosure, SSRF, and denial of service through parser exploitation.

File Upload

Insecure file upload testing including content-type bypass, polyglot files, and execution context exploitation.

Mass Assignment

Parameter binding vulnerabilities enabling privilege escalation and unauthorized field modification.

Open Redirect

URL redirection testing for phishing, OAuth bypass, and SSRF chaining opportunities.

Information Disclosure

Sensitive data exposure through error messages, debug endpoints, and metadata leakage.

Race Conditions

Time-of-check to time-of-use vulnerabilities in financial transactions, quotas, and resource allocation.

Authorization (Function-Level)

Broken function-level authorization testing for vertical and horizontal privilege escalation.

Subdomain Takeover

DNS record exploitation for subdomain hijacking across cloud providers and CDN services.

​
Skill Structure

Each vulnerability skill includes:

​
Attack Surface

Identification of vulnerable components, integration points, and input locations specific to the vulnerability class.

​
Key Vulnerabilities

Core exploitation patterns with practical examples and variations. Includes framework-specific techniques and modern variants.

​
Bypass Techniques

Methods to evade common defenses including WAFs, input filters, and validation mechanisms.

​
Testing Methodology

Step-by-step process for identifying, confirming, and validating the vulnerability across different contexts.

​
Validation Requirements

Criteria for confirming true positives with reproducible evidence and impact demonstration.

​
False Positives

Common scenarios that appear vulnerable but are properly mitigated, helping agents avoid reporting noise.

​
Example: SQL Injection Skill

The SQL injection skill provides comprehensive coverage including:
  • Database-specific primitives for MySQL, PostgreSQL, MSSQL, and Oracle
  • Detection channels: error-based, boolean-based, time-based, and out-of-band (OAST)
  • Modern techniques: JSON/JSONB operators, ORM bypass, query builder exploitation
  • Bypass methods: whitespace alternatives, keyword splitting, encoding tricks
  • Validation: reliable oracle demonstration, metadata extraction, defense-in-depth bypass
Agents using this skill can test SQL injection across different database engines, ORMs, and query contexts with advanced techniques beyond basic payloads.

​
Example: XSS Skill

The XSS skill covers:
  • Context encoding rules for HTML, attributes, JavaScript, CSS, and SVG
  • DOM XSS sources and sinks with modern framework patterns
  • Mutation XSS leveraging parser repair behavior
  • CSP bypass through script gadgets, JSONP endpoints, and import maps
  • Framework-specific sinks for React (dangerouslySetInnerHTML), Vue (v-html), Angular ($sce), Svelte ({@html})
Agents equipped with this skill understand context-specific exploitation and can bypass modern defense mechanisms.

​
When to Use Vulnerability Skills

Load vulnerability skills when:
  • Testing for specific vulnerability classes in your application
  • You need advanced exploitation techniques beyond basic payloads
  • The target uses specific frameworks or databases requiring specialized knowledge
  • Defense mechanisms require bypass techniques
  • You want comprehensive coverage including validation and false positive identification
You can combine up to 5 skills per agent, mixing vulnerability skills with framework and technology skills for specialized testing.

​
Combining Skills

Vulnerability skills work effectively with framework and technology skills: 
# Test authentication in a FastAPI application
create_agent(
    task="Test JWT authentication and IDOR vulnerabilities",
    skills="authentication_jwt,idor,fastapi"
)

# Test a Next.js application for XSS
create_agent(
    task="Test for XSS vulnerabilities in server components",
    skills="xss,nextjs"
)

# Test Supabase application for authorization issues
create_agent(
    task="Test Row Level Security implementation",
    skills="idor,business_logic,supabase"
)
By combining vulnerability skills with framework-specific knowledge, agents can apply advanced techniques in the appropriate context.

Build docs developers (and LLMs) love

Get started for freeTalk to us