#python
Agent Skills for Context Engineering offers free, open skills to master context engineering—curating AI model inputs like prompts, tools, history, and docs for top performance despite limited attention. Learn fundamentals, architectures (multi-agent, memory), tools, evaluation, and new skills like hosted agents and BDI mental states via Claude Code plugins or any platform. Examples include digital brain OS and book-writing pipelines. You gain reliable AI agents that cut errors, boost speed (up to 55% faster tasks), reduce costs, and deliver accurate results for projects.
https://github.com/muratcankoylan/Agent-Skills-for-Context-Engineering
#python
Hugging Face Skills are ready-to-use folders with instructions, scripts, and tools for AI agents to handle tasks like creating datasets, training models, running evaluations, managing jobs, and publishing papers. They work seamlessly with tools like Claude Code, OpenAI Codex, Gemini CLI, and Cursor—just install via simple commands and mention the skill in your instructions, such as "Use the HF model trainer skill." This saves you time by automating complex Hugging Face Hub operations, letting your agent execute them accurately without manual coding.
https://github.com/huggingface/skills
#python
This Claude Code Telegram Bot connects your Telegram messenger to Claude AI, letting you work with your code projects from anywhere without needing a terminal. You can chat naturally with Claude to analyze, edit, or explain your code, run tests, and manage files—all from your phone. The bot maintains conversation history for each project, so Claude remembers context between chats. It includes security features like user authentication and directory protection, plus it can handle webhooks and scheduled tasks. The main benefit is productivity on the go: you get AI-powered coding assistance instantly through a messaging app you already use daily, transforming your phone into a development tool.
https://github.com/RichardAtCT/claude-code-telegram
#python
WiFi DensePose uses WiFi signals and AI to detect human poses in real-time without cameras, tracking up to 10 people at 30 FPS with sub-50ms speed. Its Rust version boosts performance 810x faster, adds fall detection, activity tracking, and a disaster module for finding survivors under rubble via vital signs and 3D location. Install easily with `pip install wifi-densepose` for privacy-safe monitoring in homes, fitness, healthcare, or emergencies—saving lives and enhancing security without visual privacy risks.
https://github.com/ruvnet/wifi-densepose
#python
Slime is a high-performance framework for post-training large language models with reinforcement learning (RL). It connects Megatron for fast training and SGLang for data generation, powering top models like GLM-4.7, Qwen3, DeepSeek V3, and Llama 3. You get efficient, flexible RL workflows with customizable data tools, cutting training time and boosting model accuracy for research or production—saving resources while achieving breakthrough results in physics, agents, and code generation.
https://github.com/THUDM/slime
#python
Claude Code hooks let you control AI coding with 13 events like blocking dangerous tools, validating prompts, adding context, and TTS alerts for tasks. Use UV scripts for secure logging, sub-agents for team workflows (builder/validator), custom outputs, and status lines. This boosts your productivity by preventing errors, automating reviews, enabling parallel agents, and giving full observability for faster, safer shipping.
https://github.com/disler/claude-code-hooks-mastery
#python
Agent Skills are reusable packages of instructions, scripts, and resources that AI agents like Codex can automatically load to complete specific tasks efficiently and consistently. Instead of repeating the same guidance repeatedly, you write a skill once and use it everywhere, which saves time and reduces errors. You can install pre-built skills automatically or create custom ones tailored to your organization's needs, then share them across teams. This modular approach transforms general AI assistants into domain specialists, enabling them to handle complex, specialized workflows reliably without requiring you to provide the same instructions in every conversation.
https://github.com/openai/skills
#python
AI-assisted code review prompts by Linux kernel expert Chris Mason speed up Linux kernel and systemd patch reviews. Install easily with bash scripts, then use slash commands like /kreview or /kdebug in tools like Claude—they auto-load project context, break big diffs into tasks, cut token costs by 40-60%, and catch more bugs. You benefit by reviewing code 30-50% faster, saving money, reducing burnout, and improving quality without replacing human checks. Works with GPT-4, Claude, and more; pairs best with semcode for navigation.
https://github.com/masoncl/review-prompts
#python
nanochat lets you train your own GPT-2 level LLM—like a basic ChatGPT—for just $73 in 3 hours on an 8xH100 GPU node using the simple speedrun.sh script. It handles tokenization, pretraining, finetuning, evaluation, inference, and a web chat UI in minimal, hackable code. You benefit by easily building, customizing, and chatting with your personal AI on a tiny budget, learning LLM mechanics hands-on without complex setups.
https://github.com/karpathy/nanochat
#python
Kimi Code CLI is a terminal AI agent that reads/edits code, runs shell commands, searches web pages, and plans tasks autonomously. Switch to shell mode with Ctrl-X, integrate with IDEs like Zed/JetBrains via ACP, enhance Zsh, or add MCP tools. Install easily and log in with /login for Kimi Code benefits like fast 100 tokens/s speeds and high quotas. This saves you time on coding/development by automating complex workflows intelligently and securely.
https://github.com/MoonshotAI/kimi-cli
#python
The Compound Marketplace provides a Claude Code plugin that transforms your development workflow through a cycle of planning, working, reviewing, and documenting learnings. By spending 80% of effort on thorough planning and code review while only 20% on execution, you build knowledge that makes each subsequent task easier. Commands like `/workflowswork` executes them with task tracking, `/workflowscompound` documents patterns for reuse. This approach prevents technical debt accumulation, keeping your codebase maintainable and future changes straightforward.
https://github.com/EveryInc/compound-engineering-plugin
#python
Grok-1 is a powerful open-source AI model with 314 billion parameters that you can download and run on your own computer. To use it, download the model weights, install required software packages, and run a simple Python script to test it. The model uses a Mixture of Experts architecture with 64 layers and can process up to 8,192 tokens of text at once. The main benefit is that you get access to a large, capable language model under an open Apache 2.0 license, allowing you to experiment with advanced AI technology locally. However, you'll need a powerful GPU with substantial memory to run it effectively.
https://github.com/xai-org/grok-1