security-engineering

Guides authorized penetration testing, code auditing, and incident response with reproducible evidence standards.

632|86|Updated Apr 29, 2023
One-click install
npx skills add https://github.com/yaklang/yaklang --skill security-engineering-yaklang
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: security-engineering
Source: https://github.com/yaklang/yaklang/tree/main/common/ai/aid/aireact/skills/security-engineering
Command: npx skills add https://github.com/yaklang/yaklang --skill security-engineering-yaklang

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Security engagements often fail due to unclear scope, unverifiable findings, or destructive testing practices. This Skill enforces a disciplined operating doctrine for security work: authorization boundaries first, evidence over assumptions, and only reproducible vulnerabilities count as confirmed. ## Core Features & Use Cases - Mode-Based Workflows: Routes tasks into attack surface mapping, red team/black-box testing, business logic testing, white-box code audit, or blue team incident response based on intent. - Evidence Gates: Requires first-hand traffic, scripts, or runnable PoCs before any finding is marked confirmed; error messages, timing differences, and version matches alone never qualify. - Safety Constraints: Prohibits destructive actions on production data, denial of service, log tampering, and out-of-scope expansion. - Use Case: Given an authorized web application target, map its full attack surface (subdomains, ports, entry points), test each entry against a baseline, and deliver a report where every confirmed vulnerability includes reproduction steps and impact evidence. ## Quick Start Perform an authorized black-box security assessment of the target I specify, mapping the attack surface first and confirming each vulnerability with reproducible evidence.

Frequently Asked Questions about security-engineering

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run an authorized penetration test with reproducible results?▼

Define target, scope, and success criteria first, then establish a baseline request for each entry point before testing. Confirm every vulnerability with first-hand traffic or scripts that a third party can rerun, and record evidence immediately upon confirmation.

What counts as a confirmed vulnerability in security testing?▼

A vulnerability is confirmed only when it demonstrates actual impact, such as data read, privilege crossing, or state change, and can be stably reproduced by following documented steps. Error messages, timing delays, status code differences, or version fingerprint matches alone do not qualify.

How should business logic vulnerabilities be tested?▼

Business logic testing requires an authenticated session and a baseline of normal high-value workflows first. Then break trust assumptions by testing step skipping, replay, race conditions, and client-submitted value parameters, comparing state before and after each action.

Does white-box code audit require a working PoC?▼

Yes, formal white-box findings require a runnable PoC, ideally a minimal environment with a one-command verification script. Dangerous function matches without a validated source-to-sink chain are only candidates and must be marked unverified.

What actions are prohibited during security testing?▼

Prohibited actions include deleting or modifying production data, planting persistent backdoors, denial of service, clearing logs, touching real funds or inventory, and testing unauthorized third parties. All write operations must use test objects and remain reversible.