sammyboi81

LLM App Security — Security skill for Claude Code

Security community

Building LLM-powered apps safely - prompt injection, excessive agency, data leakage, output handling (OWASP LLM Top 10).

How to install LLM App Security

Installs to ~/.claude/skills/sammyboi81-bulwark-mcp-llm-app-security/SKILL.md

Terminal

mkdir -p ~/.claude/skills/sammyboi81-bulwark-mcp-llm-app-security && curl -fsSL https://raw.githubusercontent.com/sammyboi81/bulwark-mcp/HEAD/skills/llm-app-security.md -o ~/.claude/skills/sammyboi81-bulwark-mcp-llm-app-security/SKILL.md

Restart Claude Code, or start a new session, for it to be picked up.

What LLM App Security does

name: llm-app-security description: Building LLM-powered apps safely - prompt injection, excessive agency, data leakage, output handling (OWASP LLM Top 10). tags: ai, llm, core

Prompt injection is unsolved: you cannot reliably stop a model from following hostile text. Design so a fully hijacked model can do little harm.

Threat model

Direct injection (user text) and INDIRECT injection (web pages, emails, PDFs, tickets, repos, tool output, retrieved chunks, image text, file metadata)

Alternatives in Security

Deepsec — Deepsec is a security harness for finding vulnerabilities in your codebase powered by coding agents 7.8k ★
Claude Code Security Review — An AI-powered security review GitHub Action using Claude to analyze code changes for security vulnerabilities 3.9k ★
Promptmap — Security scanner for custom LLM apps 1.2k ★