Skill UI

A comprehensive skill for comparing the GAIA benchmark harness across various implementations, such as ruflo and Princeton HAL. It analyzes capability gaps (e.g., real code execution, full question sets, browser automation) and provides a detailed improvement roadmap for enhancing AI evaluation benchmarks.

This skill runs an automated evolution process designed to systematically optimize the seven core policy surfaces of an AI agent harness (e.g., planner, contextBuilder, reviewer). By mutating these policies and scoring the variants in a sandbox, it identifies crucial performance improvements without requiring expensive foundation model retraining. It is ideal for empirical configuration tuning and establishing optimized baselines.

A powerful tool designed to scaffold sophisticated, custom AI agent harnesses using the metaharness CLI. It supports various templates (e.g., coding, devops, marketing) and AI hosts (e.g., Claude, Codex). Crucially, it operates with maximum safety, defaulting to a dry-run mode and refusing to write to the calling repository root, ensuring controlled and predictable development workflows.

A comprehensive framework for optimizing inventory levels and planning replenishment in multi-location retail environments. It covers advanced demand forecasting techniques (e.g., Holt-Winters, ML models), calculating safety stock based on variability, estimating promotional uplift, and generating actionable purchase orders while minimizing stockouts and excess inventory. Ideal for professionals managing complex SKU portfolios across various supply chain systems.

This tool provides state-of-the-art semantic memory search across various knowledge domains. It supports multiple advanced retrieval strategies, including hybrid (sparse+dense) searching and Graph RAG for multi-hop reasoning. Key features include Mean Margin Ranking (MMR) for result diversity, recency weighting for timely information, and structured namespace filtering, ensuring highly relevant and context-aware results.

This resource is designed to document and guide the implementation of various architectural patterns and best practices when developing backend applications using the NestJS framework. It helps developers structure their code for maximum scalability and maintainability.

Implements the DISTILL and CONSOLIDATE phases of the intelligence pipeline. This tool enables the training and consolidation of learned neural patterns. It supports various adaptation methods, including single-domain SONA for real-time learning and multi-domain MicroLoRA for robust, long-term knowledge storage, preventing catastrophic forgetting.

This tool aggregates various system metrics—including counters, gauges, and histograms—from the observability namespace. It calculates baselines, flags significant deviations (anomalies) based on statistical standards, and provides a comprehensive system health report. Use it to get a detailed snapshot of system performance, monitor for degradation, and ensure overall system stability.

This tool detects and flags Personally Identifiable Information (PII) across text, code, and configuration files. It is essential to use this before committing code, writing logs, or generating model responses to prevent accidental exposure of sensitive data. It identifies various types of PII, including emails, phone numbers, SSNs, credit card numbers, API keys, and passwords, offering quick checks, detailed scans, and full context-aware analysis.

This tool provides comprehensive safety scanning for untrusted inputs, such as user submissions, API payloads, and webhook data. It actively detects various security threats, including prompt injection, jailbreak attempts, instruction override attacks, and unsafe content. Beyond basic checks, it offers deep analysis, full multi-layer scanning, and even capabilities to train defensive models using detected threats. Essential for secure AI application development.

This tool assesses whether coding agents adhere to defined skills, rules, or agent definitions. It automatically generates scenarios with varying prompt strictness, executes the agent, and generates a comprehensive report detailing the compliance rate of tool calls against predefined workflow specifications using advanced LLM classification.

This tool initializes a structured, anti-drift multi-agent swarm environment. It is designed for coordinating complex, multi-file software development tasks, such as implementing new features, performing cross-module refactoring, or conducting security audits, requiring coordination among three or more specialized agents. It supports various topologies (hierarchical, mesh, etc.) to manage safe and parallel workstreams.