Skill UI

Guides users through the entire end-to-end process of running the GAIA benchmark. It covers environment validation, cost estimation, model execution, and generating a signed, HAL-compatible submission package for leaderboards. This ensures high integrity and reproducibility for model performance evaluation.

This advanced workflow is designed to generate comprehensive, fact-checked research reports on any complex topic. It systematically performs fan-out web searches across multiple defined angles, fetches multiple sources, and employs a rigorous adversarial verification system (3-vote) to validate claims. The final output is a highly cited, synthesized report, ideal for academic or critical business research.

A comprehensive framework for creating and executing intelligent, goal-oriented action plans (GOAP). This system analyzes preconditions, optimizes costs, and provides adaptive replanning capabilities. It is designed for complex objectives that require multiple, dependent steps, ensuring robust task completion even when unexpected state changes occur.

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.

This tool provides deep inspection and auditing capabilities for GEPA (genetic-evolution prompt-adaptation) genomes by utilizing the internal `@metaharness/darwin/gepa` library. It is crucial for advanced AI development pipelines, allowing users to validate the structural integrity of genomes, render the full system prompt derived from a genome, or classify specific failure modes found in an execution transcript. Ideal for CI/CD gating and debugging complex LLM applications.

This tool executes a GEPA learning cycle, automating the evolution and optimization of policy prompts against a scored task corpus. Instead of relying on manual prompt iteration, it measures performance improvements, identifying the best working 'harness genome.' It is ideal for measuring sustainable improvements in task performance and supports a free, cost-estimating dry-run mode before committing to actual model calls.

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.

This tool executes the advanced `metaharness/darwin` security benchmark (ADR-155). It rigorously evaluates the security detection capabilities of AI systems against a controlled ground-truth corpus, providing critical metrics like TPR, FPR, and patch-pass rates. It compares the champion model against multiple baselines, offering empirical data crucial for iterative security improvement and drift detection.

A sophisticated system designed to automatically select the optimal AI agent and model tier (e.g., Haiku, Sonnet, Opus) for any non-trivial task. It replaces manual decision-making by integrating learned patterns, predictive modeling, and a robust 3-tier routing mechanism. The process is fully auditable, providing a detailed routing rationale to ensure transparency and reliable execution.

This skill facilitates the crucial process of sharing learned intelligent patterns across isolated projects or machines using the InterPlanetary File System (IPFS). It allows users to publish a project's accumulated knowledge base to a decentralized network and later fetch, apply, and merge these patterns into new or existing projects. This is invaluable for bootstrapping new applications with prior knowledge or consolidating learnings across a large monorepo.

This skill guides the architectural design of an Itô Data Atlas agent for advanced quantitative research. It facilitates market discovery, constructs candidate prediction-market baskets, drafts necessary parameter changes, and integrates human-in-the-loop review. It is strictly for workflow and architectural planning, and does not execute live trading orders.

Provides comprehensive configuration management for local LLM inference using RuVLLM. Users can select models, perform task-specific fine-tuning with MicroLoRA adapters, or implement continuous, real-time adaptation using SONA. This tool guides the entire advanced LLM workflow, covering status checks, configuration generation, and advanced adaptation loops.