Skill UI

This tool converts local source audio recordings into a specified, authorized target voice. It manages the complex workflow, including checking source and target voice rights, handling linguistic specifics (locale, accent), and performing previews before generating final assets. It is essential for professional narration, podcasting, and content localization while adhering to strict copyright and identity policies.

A comprehensive voiceover studio that transforms written text into high-quality, professional spoken audio. It supports generating content for podcasts, IVR prompts, educational videos, and product demos. The system handles complex requirements including target locale, specific accents (e.g., British, Mandarin), desired emotion, and pacing, ensuring the resulting audio artifact is natural and contextually accurate. Ideal for content creators and developers needing reliable, multi-lingual narration.

A specialized quality gate skill designed to review complex, multi-source outputs generated by the CrossFrame framework. It rigorously checks reasoning fidelity, evidence boundaries, source anchors, and structural continuity. Use this when advanced verification of logical rigor and source integrity is required, ensuring outputs meet strict internal protocols and avoid common structural failures.

This skill uses Python and boto3 to analyze AWS CloudTrail logs. It establishes statistical baselines of normal API activity and detects various anomalies, such as unusual event sources, first-time API usage, geographic IP deviations, and high-frequency calls. This is critical for identifying credential compromise, privilege escalation, and unauthorized resource access in a security incident.

Automated system designed to monitor paste sites (e.g., Pastebin, GitHub Gists) for leaked credentials, API keys, and sensitive data dumps. It uses advanced scraping, keyword matching, and regex patterns to provide early detection of data breaches and security incidents.

This skill outlines the implementation of advanced runtime defenses for production Large Language Model (LLM) applications. It covers using multiple industry-leading guardrail systems—including Llama Guard, NeMo Guardrails, and LLM Guard—to detect and block adversarial attacks such as prompt injection, jailbreaks, and toxic content. It provides a comprehensive defense-in-depth strategy for ensuring model safety and output integrity in real-world deployments.

This skill provides comprehensive tools to detect data poisoning and model backdooring attacks throughout the machine learning lifecycle. It checks data provenance, detects label-flip anomalies (using Cleanlab), and analyzes model activations and spectra (using ART) to identify malicious triggers or compromised weights, ensuring ML supply chain integrity.

A comprehensive security tool designed to detect and defend against subtle prompt injection attacks hidden within untrusted multi-modal content, such as web pages, PDFs, and images. It scans for obfuscated payloads (e.g., hidden CSS, metadata, zero-width characters) by combining content normalization, heuristic analysis, and dedicated ML detectors, ensuring that LLM agents only consume sanitized, trustworthy input.

This skill provides methods and frameworks to detect sophisticated AI attacks, including model stealing, membership inference, and model inversion. It focuses on monitoring query patterns, analyzing confidence exposure, and implementing defensive controls on inference APIs to protect intellectual property and data privacy.

This skill utilizes NVIDIA's open-source garak framework to conduct comprehensive red-teaming assessments on Large Language Models (LLMs). It tests for critical vulnerabilities such as prompt injection, jailbreaks, data leakage, and toxic content generation by sending thousands of adversarial probes. Ideal for pre-deployment security validation, API guardrail testing, and generating defensible evidence for AI risk assessments.

This skill provides a comprehensive defense-in-depth framework for securing autonomous AI agents. It implements least-privilege tool allowlisting, identity binding, and human-in-the-loop controls to mitigate risks like prompt injection and tool poisoning. Use this when building agents that interact with sensitive systems (e.g., payments, file writes) to ensure governance, auditability, and compliance.

The Anti-Deception Harness detects manipulative user prompts, such as manufactured urgency, pressure to validate, or undue reliance on authority. It calls an external tool to provide a structured scaffold detailing integrity procedures and counter-evidence, ensuring responses are honest and robust, preventing compromised or sycophantic answers. Ideal for high-stakes decision support.