How do I install the Security skill?

Run npx skillmds@latest add jylhis/security in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Security safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Security?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Security free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Security?

Jylhis (@jylhis) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Security by jylhis — Agent Skill

Name: Security
Author: Jylhis

Security

Use for code security review and safe alternatives across language ecosystems — Python (subprocess command injection, pickle / yaml.load deserialization, SSRF, SQL injection, path traversal, secrets, bcrypt/argon2, cryptography, TLS verify); TypeScript / Node (child_process exec/shell, prototype pollution, XSS via DOMPurify, SSRF, secrets handling, Zod validation, npm audit, Helmet, CSP); JVM (JNDI / Log4Shell, Java serialization, XXE, SSRF, SQL injection, secrets, crypto APIs). Also covers framework-specific secure-by-default guidance (Django, FastAPI, Flask, Express, React, Next.js, Vue, Go), repository threat modeling, security ownership mapping, and CodeQL static analysis. Read the matching reference before reviewing untrusted-input handlers.

Jylhis 53e0d75 20 files · 449.3 KB Updated Sep 17, 2026

File contents

Security skill index

Pick the topic and read its reference before auditing or writing any code that handles untrusted input or external services.

Stack	When to read	Reference
Python	subprocess command injection, pickle / yaml.load, SSRF, SQL injection, path traversal, secrets, bcrypt / argon2, cryptography, TLS verify	`references/python.md`
TypeScript / Node	child_process exec, prototype pollution, XSS, SSRF, secrets, Zod validation, npm audit, Helmet, CSP / HSTS	`references/typescript.md`
JVM (Java / Kotlin)	JNDI / Log4Shell-class, Java serialization, XXE, SSRF, SQL injection, secrets, crypto APIs	`references/jvm.md`
Best practices by framework	writing secure-by-default code or producing a prioritized vulnerability report; per-framework guidance under `references/best-practices/` (Django, FastAPI, Flask, Express, React, Next.js, Vue, jQuery, Go backend)	`references/best-practices.md`
Threat modeling	building an evidence-anchored AppSec threat model of a repo: trust boundaries, assets, entry points, abuse cases, prioritized risks	`references/threat-model.md`
Ownership map	mapping which team / owner is responsible for each security-relevant component	`references/ownership-map.md`
Static analysis (CodeQL)	running or authoring CodeQL queries for Python, JS/TS, Go, Java/Kotlin, C/C++, C#, Ruby, Swift	`references/static-analysis-codeql.md`

Common rules across all stacks:

Validate untrusted input at system boundaries (HTTP handlers, message queues, file uploads, env-var parsers).
Never pass untrusted strings to a shell — use argv-array forms (subprocess.run([...]), child_process.execFile(...)).
Block private / metadata IP ranges (e.g. 169.254.169.254) on any outbound HTTP that takes a user-supplied URL.
Treat secrets as inputs from a secret manager; never commit .env or hard-code credentials.

After reading the reference, follow its guidance for the task.

---name: securitydescription: Use for code security review and safe alternatives across language ecosystems — Python (subprocess command injection, pickle / yaml.load deserialization, SSRF, SQL injection, path traversal, secrets, bcrypt/argon2, cryptography, TLS verify); TypeScript / Node (child_process exec/shell, prototype pollution, XSS via DOMPurify, SSRF, secrets handling, Zod validation, npm audit, Helmet, CSP); JVM (JNDI / Log4Shell, Java serialization, XXE, SSRF, SQL injection, secrets, crypto APIs). Also covers framework-specific secure-by-default guidance (Django, FastAPI, Flask, Express, React, Next.js, Vue, Go), repository threat modeling, security ownership mapping, and CodeQL static analysis. Read the matching reference before reviewing untrusted-input handlers.---# Security skill indexPick the topic and read its reference before auditing or writing anycode that handles untrusted input or external services.| Stack | When to read | Reference ||---|---|---|| Python | subprocess command injection, pickle / yaml.load, SSRF, SQL injection, path traversal, secrets, bcrypt / argon2, cryptography, TLS verify | `references/python.md` || TypeScript / Node | child_process exec, prototype pollution, XSS, SSRF, secrets, Zod validation, npm audit, Helmet, CSP / HSTS | `references/typescript.md` || JVM (Java / Kotlin) | JNDI / Log4Shell-class, Java serialization, XXE, SSRF, SQL injection, secrets, crypto APIs | `references/jvm.md` || Best practices by framework | writing secure-by-default code or producing a prioritized vulnerability report; per-framework guidance under `references/best-practices/` (Django, FastAPI, Flask, Express, React, Next.js, Vue, jQuery, Go backend) | `references/best-practices.md` || Threat modeling | building an evidence-anchored AppSec threat model of a repo: trust boundaries, assets, entry points, abuse cases, prioritized risks | `references/threat-model.md` || Ownership map | mapping which team / owner is responsible for each security-relevant component | `references/ownership-map.md` || Static analysis (CodeQL) | running or authoring CodeQL queries for Python, JS/TS, Go, Java/Kotlin, C/C++, C#, Ruby, Swift | `references/static-analysis-codeql.md` |Common rules across all stacks:- Validate untrusted input at system boundaries (HTTP handlers, message queues, file uploads, env-var parsers).- Never pass untrusted strings to a shell — use argv-array forms (`subprocess.run([...])`, `child_process.execFile(...)`).- Block private / metadata IP ranges (e.g. `169.254.169.254`) on any outbound HTTP that takes a user-supplied URL.- Treat secrets as inputs from a secret manager; never commit `.env` or hard-code credentials.After reading the reference, follow its guidance for the task.

Frequently asked questions