Skill UI

A comprehensive toolkit for designing, architecting, and evaluating complex multi-agent systems. It provides structured patterns—including Supervisor, Swarm, and Pipeline—for system architecture, defines agent roles (Coordinator, Specialist), and covers best practices for communication (Event Sourcing, Message Passing), tool design, and safety guardrails, ensuring robust and scalable AI workflows.

Guides building reliable autonomous agents by covering agent loops, goal decomposition, reflection, and tooling patterns while stressing guardrails, human review, and auditability to curb compounding failure rates in production.

LlamaGuard is Meta’s 7–8B safety-specialized LLM that filters both prompts and responses by classifying six threat categories, enabling fast inference via vLLM/SageMaker and integration into NeMo Guardrails for end-to-end moderation.

NVIDIA’s NeMo Guardrails adds runtime safety rails to LLM apps, covering jailbreak detection, hallucination and fact-checking, PII filtering, and toxicity checks via Colang 2.0 rules and integrations for production use on T4 GPUs.

This guide implements comprehensive guardrails for Perplexity Sonar API. It covers crucial aspects like query content moderation, PII sanitization, enforcing user usage quotas, selecting appropriate models based on user tiers, and scoring citation reliability to ensure safe, ethical, and cost-effective AI usage.

Designs, optimizes, and evaluates prompts for LLM applications by defining success criteria, choosing chain-of-thought or few-shot formats, testing diverse cases, enforcing structured schema outputs, and documenting guardrails so deployments stay accurate and efficient.

This skill provides implementation patterns for policy guardrails when utilizing Retell AI for voice agents and phone call automation. It is designed to help ensure that AI interactions in a telephony context adhere strictly to predefined business rules and safety policies, preventing unauthorized actions or deviations from the intended conversational flow.

Builds NeMo Guardrails-based input/output validation pipelines to block prompt injection, redact PII, enforce content policies, and verify structured outputs so LLM-powered services stay safe before downstream use.

Switches the AI persona into a Chinese-style mama nagging narrator that sticks to core behavior guardrails while delivering increasingly intense motivational scripts; useful when you need a relentless reminder that you must research, verify, and finish tasks before asking for help.

This guide details how to implement robust content safety guardrails for applications powered by Claude. It covers essential techniques such as system prompt hardening, input validation to detect injection patterns and length limits, and output validation to prevent system prompt leakage. It ensures compliance with Anthropic's policies, making the AI integration secure and reliable.

This comprehensive skill provides multi-layered guardrails for integrating with LLMs like Claude. It enforces security and compliance by implementing input validation (PII detection, length checks), system prompt restriction (scope control, safety rules), output filtering (preventing sensitive data leakage), cost governance, and model access policies. Ideal for building secure, robust, and governed AI applications.