Detect prompt injection attempts in AI and LLM applications.
npx skills add https://github.com/ricardoroche/ricardos-claude-code --skill ai-security
Please help me install this Agent Skill. Skill: ai-security Source: https://github.com/ricardoroche/ricardos-claude-code/tree/main/.claude/skills/ai-security Command: npx skills add https://github.com/ricardoroche/ricardos-claude-code --skill ai-security
This Skill addresses critical security vulnerabilities in AI and LLM applications, such as prompt injection, PII leakage, and generation of unsafe content. It provides robust patterns to protect your applications and users, ensuring compliance and trust.
SecurePromptBuilder to detect injection, redact PII, and filter the LLM's response for harmful content, ensuring a safe interaction.Implement prompt injection detection for user inputs in my LLM application.
High-intent search queries and answers about installing and using this skill.
Prompt injection detection identifies malicious inputs designed to manipulate LLM behavior. This Skill implements rule-based detectors that scan user queries for injection patterns, flagging suspicious instructions before they reach your model. Apply it at API endpoints and orchestration layers to block threats early.
Yes. PII redaction automatically removes sensitive personally identifiable information from prompts and unredacts outputs when safe. This prevents accidental data leakage to external LLMs and ensures compliance with privacy regulations across prompt handling and response generation.
Output content filtering screens LLM responses for violence, hate speech, sexual content, and illegal activities. This Skill applies configurable sensitivity rules to generated text, catching unsafe outputs before they reach users and maintaining application safety standards.
Secure prompt construction combines clear boundaries and built-in validation through a SecurePromptBuilder pattern. It detects injection attempts, redacts sensitive data, and filters responses in a single workflow, reducing manual security configuration across your application.
Yes. The Skill spans API endpoints, agents, and orchestration layers with structured diagnostics and configurable sensitivity. It integrates with security engineers and ML-system architectures, supporting both simple endpoints and complex multi-component deployments.
This Skill depends on Pydantic for data validation and structured configuration. Pydantic provides the schema enforcement needed for rule-based detection and diagnostic output, enabling integration with existing Python LLM frameworks.
Renzo-TognellaDefine and validate Pydantic data models for Python schemas.
hyperb1issAutomate security reviews, threat modeling, and incident response guidance.
ejolietDesign prompts, evaluate LLM outputs, and build agentic workflows.
truefoundryConfigure content safety guardrails for AI Gateway endpoints.