A deterministic, zero-dependency Python tool that checks whether apparent evaluation anomalies survive simple alternative explanations

When chain of thought monitors trustworthy?

Technical AI safety Puzzle

The Feature Hunter Agent for LLM Interpretability

Rust based CLI tool for estimating success rates when using LLM judges for evaluation.

Jane street dormant puzzle trail runs

Rust Agentic Skills : It transforms general-purpose LLMs into a multi-agentic Rust engineering team via a token-efficient Brain-Tool-Context setup that dynamically loads tools and documentation, ad…

Open-source, Agentic MLOps Skill for local, privacy-first SLM fine-tuning, using Unsloth for efficient QLoRA training and Ollama/lmstudio for deployment.

Unofficial Python SDK for OpenResponses Spec examples.

Universal Shopping AI Assistant using Universal Commerce Protocol (UCP)

Evaluation Agent that evaluates an agent that runs tasks in virtual EHR workflows.

Medical Agent Benchmark - FHIR MCP Server

Institutional-grade async news arbitrage engine for NSE/BSE with WAF bypass and high-frequency execution

Paper2Agent with Skills is featuring a Robust Adaptive Skill-Accretion (R-ASA) engine. It transforms research papers into active agents that dynamically synthesize, verify, and memorize executable …