Skill UI

Stable Baselines3 is a robust, PyTorch-based library providing production-ready implementations of core reinforcement learning algorithms (PPO, SAC, DQN, etc.). It facilitates the training of single-agent RL models, supports the creation of custom Gymnasium environments, and handles advanced parallel training via vectorized environments. It is ideal for rapid prototyping and rigorous RL experiments.

This skill deploys advanced anomaly detection systems for Operational Technology (OT) and Industrial Control Systems (ICS). It builds multi-dimensional baselines by analyzing SCADA communications, network traffic (Modbus, DNP3, OPC UA), and timing patterns. It detects deviations, rogue devices, and potential intrusions by profiling normal behavior, complementing signature-based IDS.

Utilizes User and Entity Behavior Analytics (UEBA), statistical baselines, and machine learning models to monitor authentication logs. This skill identifies sophisticated security threats such as impossible travel, credential stuffing, brute force attacks, and compromised accounts in real-time. Essential for security operations and identity access management.

This skill acts as an advanced extension layer, implementing a self-evolution protocol, maintaining session state continuity (Compaction), and providing KPI-style performance summaries. It allows users to simulate continuous professional development, track metrics, and master complex role-playing and interaction skills across multiple sessions, ensuring performance baselines are continuously adjusted and internalized.

Stable Baselines3 is a robust, PyTorch-based library providing reliable implementations of state-of-the-art reinforcement learning algorithms (PPO, SAC, DQN, etc.). It is designed for standard RL experiments, quick prototyping, and developing complex single-agent systems. The library supports training agents, creating custom Gymnasium environments, and running vectorized simulations for accelerated learning.

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.

This skill runs comprehensive benchmarks against structural and adversarial corpora. It measures the performance and cost efficiency of a booster model compared to leading LLM baselines (Gemini, Anthropic). Results are persisted in verifiable JSON reports, crucial for auditing claims and verifying booster win rates before publishing.

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 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.