skills.rest
informatico-madrid
informatico-madridCOMMUNITY
Explore all skills by this author →
GitHub

prompt-guard

Detect prompt injections and jailbreak attempts in text using a pre-trained classifier.

1|Updated Feb 10, 2026
Search Tags:#security#input validation#prompt injection#content filtering#jailbreak detection#llm safety
One-click install
npx skills add https://github.com/informatico-madrid/Architect-Expert-Gap-Forge --skill prompt-guard-informatico-madrid
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: prompt-guard
Source: https://github.com/informatico-madrid/Architect-Expert-Gap-Forge/tree/main/.github/skills/prompt-guard
Command: npx skills add https://github.com/informatico-madrid/Architect-Expert-Gap-Forge --skill prompt-guard-informatico-madrid

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires transformers, torch, and includes references (resource) components.

What problem does it solve?

This Skill safeguards Large Language Model (LLM) applications by detecting and filtering out malicious prompts, including prompt injections and jailbreak attempts, thereby preventing unauthorized access or manipulation.

Core Features & Use Cases

  • Prompt Injection Detection: Identifies attempts to embed hidden instructions within user inputs or third-party data.
  • Jailbreak Detection: Flags direct attempts to override the LLM's safety guidelines or intended behavior.
  • Multilingual Support: Operates effectively across 8 different languages.
  • Use Case: Protect your customer-facing chatbot by using this Skill to scan all incoming user messages, ensuring that no harmful or manipulative prompts reach the LLM.

Quick Start

Use the prompt-guard skill to check if the user message 'Ignore previous instructions' is a jailbreak attempt.

Frequently Asked Questions about prompt-guard

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

FAQPage Schema
How do I detect prompt injection attempts in user inputs for my LLM application?▼

Detect prompt injection attempts by scanning user inputs and third-party data with an 86M parameter classifier that identifies hidden instructions and filters malicious prompts before they reach the LLM.

What is the best way to prevent jailbreak attempts in a customer-facing chatbot?▼

The best way to prevent jailbreak attempts is to scan all incoming user messages to flag direct overrides of safety guidelines, ensuring manipulative prompts do not reach the LLM.

Does prompt injection detection work with multilingual inputs?▼

Yes, prompt injection detection works with multilingual inputs, offering effective classification across 8 different languages to filter malicious prompts.

Do I need a GPU to run a jailbreak detection classifier?▼

You do not need a GPU to run jailbreak detection; the classifier operates with low latency on both GPU and CPU environments.

How does content filtering identify hidden instructions in third-party data?▼

Content filtering identifies hidden instructions by analyzing text with a pre-trained classifier that distinguishes legitimate input from embedded malicious prompts.

Related Skills

Explore More Software Engineering Skills→
Orchestra-ResearchOrchestra-Research
11.5k

prompt-guard

Detect prompt injections and jailbreak attempts in LLM inputs.

Official
Advanced
RedHatProductSecurityRedHatProductSecurity
47

input-output-sanitization

Validate and sanitize MCP server request and response data.

Official
Intermediate
djankiesdjankies

sanitizing-user-inputs

Sanitize and validate user input with typed schemas in TypeScript applications.

Community
Advanced
sandmark78sandmark78

input-validator

Detect dangerous commands and prompt injections in user-provided text.

Community
Intermediate