Why does this skill require a threat model first?

A threat model provides the necessary context and priority to prevent the AI from generating generic, irrelevant security lists. It focuses the analysis on high-impact, reachable code paths.

Can I use this for general code quality reviews?

No, this skill is specifically optimized for security vulnerability hunting and requires a different level of rigor than standard linting or refactoring tasks.

What is depth-first analysis in this context?

It means fully investigating one attack surface or code path to prove or disprove exploitability before moving to the next, rather than scanning the whole codebase superficially.

How does this skill prevent AI hallucinations in security reports?

It mandates concrete evidence, such as specific file and line references for data flow traces and actual exploit payloads, refusing to report findings that are merely theoretical.

Targeted Vulnerability Analysis

Name: Targeted Vulnerability Analysis
Author: kylejryan

bykylejryan

•

安全与测试

Performs rigorous, depth-first security audits by tracing data flows and proving exploitability based on established threat models.

Targeted Vulnerability Analysis is a specialized engineering skill designed to eliminate the noise of theoretical security scans and breadth-first hallucinations. By enforcing a depth-first methodology, it moves beyond simple pattern-matching to provide verified source-to-sink code traces and concrete exploit payloads. This skill requires an existing threat model to prioritize attack surfaces, ensuring that analysis remains focused on high-impact areas while documenting both confirmed vulnerabilities and investigated-but-secure paths for total transparency in the auditing process.

主要功能

01Concrete payload construction for proof-of-concept findings

02Comprehensive reporting of both confirmed vulnerabilities and safe paths

03Depth-first analysis methodology for high-fidelity security reviews

041 GitHub stars

05Suppression of breadth-first pattern-matching and hallucinations

06Source-to-sink data flow tracing to prove exploitability

使用场景

01Conducting deep-dive security audits of mission-critical application code

02Reducing false positives in automated AI-driven security review workflows

03Verifying potential vulnerabilities identified during threat modeling sessions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add kylejryan/better-code targeted-vuln-analysis

For use in Claude.ai and ChatGPT

主要功能

01Concrete payload construction for proof-of-concept findings

02Comprehensive reporting of both confirmed vulnerabilities and safe paths

03Depth-first analysis methodology for high-fidelity security reviews

041 GitHub stars

05Suppression of breadth-first pattern-matching and hallucinations

06Source-to-sink data flow tracing to prove exploitability

使用场景

01Conducting deep-dive security audits of mission-critical application code

02Reducing false positives in automated AI-driven security review workflows

03Verifying potential vulnerabilities identified during threat modeling sessions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add kylejryan/better-code targeted-vuln-analysis

For use in Claude.ai and ChatGPT