Documentation for Hermes Agent, a self-improving AI agent built by Nous Research. It describes a learning loop that creates skills from experience, improves them during use and remembers across sessions. The docs cover installation, context files, the memory and skills systems, voice mode, and connecting external tool servers via MCP.
Tensor Trust is a prompt injection attack and defense game that invites users to fool AI language models and climb a leaderboard. The site is part of a research project studying prompt injection vulnerabilities in AI systems, using crowdsourced submissions to help researchers build more reliable AI.
goose is an open source AI agent that runs on the user's machine. It is available as a desktop app for macOS, Linux and Windows, a CLI, and an API built in Rust. The agent is general-purpose, covering code, workflows, research, writing and automation. It connects to 70+ extensions via the Model Context Protocol and works with 15+ LLM providers, including API keys or existing subscriptions.
DeepLearning.ai is an educational platform led by Andrew Ng that offers online AI courses, specializations, and certifications for people learning to use and build artificial intelligence. The platform serves over 7 million learners and covers topics including AI Python, prompt engineering, generative AI, and machine learning.
A guide to prompt engineering that collects papers, advanced prompting techniques, learning guides, model-specific guides, lectures, references, new LLM capabilities, and tools related to prompt engineering. It describes prompt engineering as a discipline for developing and optimizing prompts to use language models across applications and research topics, and offers courses on prompting and building AI agents.
VoxelBench provides machine-learning models, research, training resources, or evaluation tools.
SimpleBench is a multiple-choice text benchmark for LLMs with over 200 questions covering spatio-temporal reasoning, social intelligence and trick questions. Its report states a non-specialized human baseline of 83.7 percent from a small sample of nine participants.
Community notes
Be the first to share how this project worked for you.
Sign in with a verified account to leave a review or join the discussion.
Sign in to reviewKeep exploring
Tensor Trust is a prompt injection attack and defense game that invites users to fool AI language models and climb a leaderboard. The site is part of a research project studying prompt injection vulnerabilities in AI systems, using crowdsourced submissions to help researchers build more reliable AI.
goose is an open source AI agent that runs on the user's machine. It is available as a desktop app for macOS, Linux and Windows, a CLI, and an API built in Rust. The agent is general-purpose, covering code, workflows, research, writing and automation. It connects to 70+ extensions via the Model Context Protocol and works with 15+ LLM providers, including API keys or existing subscriptions.
DeepLearning.ai is an educational platform led by Andrew Ng that offers online AI courses, specializations, and certifications for people learning to use and build artificial intelligence. The platform serves over 7 million learners and covers topics including AI Python, prompt engineering, generative AI, and machine learning.
A guide to prompt engineering that collects papers, advanced prompting techniques, learning guides, model-specific guides, lectures, references, new LLM capabilities, and tools related to prompt engineering. It describes prompt engineering as a discipline for developing and optimizing prompts to use language models across applications and research topics, and offers courses on prompting and building AI agents.
VoxelBench provides machine-learning models, research, training resources, or evaluation tools.
SimpleBench is a multiple-choice text benchmark for LLMs with over 200 questions covering spatio-temporal reasoning, social intelligence and trick questions. Its report states a non-specialized human baseline of 83.7 percent from a small sample of nine participants.
