defending-llms-with-guardrails

80 / 100 · Low risk

Fine to try. Some of what it can do is expected: a security skill usually runs scanners and checks websites.

By mukul975/Anthropic-Cybersecurity-Skills (GitHub) · security · 35,906 stars · scanned at commit 54a7988

Deploys Llama Guard 3 safety classification, NeMo Guardrails programmable dialogue rails, and LLM Guard input/output scanner pipelines as complementary runtime defenses that inspect and constrain LLM prompts and responses. Use when adding a production runtime safety layer to an LLM, RAG, or agent application to block jailbreaks, prompt injection (OWASP LLM01), toxic content, or sensitive-data leakage before it reaches or leaves the model.

Works with any agent that supports SKILL.md skills.

What this skill can do

Downloads and installs other software

The author does not mention this.

Found in 5 places.

skills/defending-llms-with-guardrails/SKILL.md, line 53python -m pip install llm-guard
skills/defending-llms-with-guardrails/SKILL.md, line 56python -m pip install nemoguardrails
skills/defending-llms-with-guardrails/SKILL.md, line 59python -m pip install "transformers>=4.43" torch accelerate huggingface_hub

Show 2 more

skills/defending-llms-with-guardrails/scripts/agent.py, line 44sys.exit("[!] llm-guard not installed. Run: pip install llm-guard")
skills/defending-llms-with-guardrails/scripts/agent.py, line 74sys.exit("[!] transformers/torch not installed. Run: pip install 'transformers>=4.43' torch accelerate")

Runs commands on your computer

The author does not mention this. Expected for a security skill.

Found in 2 places.

skills/defending-llms-with-guardrails/SKILL.md, line 53python -m pip install llm-guard
skills/defending-llms-with-guardrails/SKILL.md, line 258python scripts/agent.py llmguard --input payloads.jsonl --report report.json

Install

Paste this into your terminal. It downloads exactly the version we scanned, not whatever is newest.

if [ -e ~/.agents/skills/'defending-llms-with-guardrails' ] || [ -L ~/.agents/skills/'defending-llms-with-guardrails' ]; then echo 'A skill folder named defending-llms-with-guardrails already exists — not replacing it' >&2; false; else d=$(mktemp -d) && git clone --quiet --filter=blob:none --no-checkout 'https://github.com/mukul975/Anthropic-Cybersecurity-Skills' "$d" && git -C "$d" checkout --quiet '54a798831d2266a3ca61ce68a7acb80b81160d57' -- 'skills/defending-llms-with-guardrails' && mkdir -p ~/.agents/skills && cp -R "$d"/'skills/defending-llms-with-guardrails' ~/.agents/skills/'defending-llms-with-guardrails' && rm -rf "$d"; fi

Codex reads ~/.agents/skills directly. For Claude Code, Cursor, Copilot, Gemini CLI, Antigravity or OpenCode, also link it. For Claude Code: ln -s ~/.agents/skills/'defending-llms-with-guardrails' ~/.claude/skills/'defending-llms-with-guardrails' (other agents: ~/.cursor/skills, ~/.copilot/skills, ~/.gemini/skills, ~/.gemini/config/skills, ~/.config/opencode/skills). Or install through the finder skill, which does both.

Badge for your README

The badge for mukul975/Anthropic-Cybersecurity-Skills shows the score of its lowest-scoring skill.

SkillTrust badge for mukul975/Anthropic-Cybersecurity-Skills

[![SkillTrust](https://zalavijaysinh.github.io/skilltrust/badges/mukul975/Anthropic-Cybersecurity-Skills.svg)](https://zalavijaysinh.github.io/skilltrust/repos/mukul975/Anthropic-Cybersecurity-Skills/)

Score history

Date	Version	Score
2026-11-08	`54a7988`	80
2026-11-07	`54a7988`	80