📦 poorna-repos/opus-fable-mode

Make Claude Opus 4.8 Act Like Fable 5

Ever feel like Claude Opus is a bit too talkative or anxious when you use it for coding? After Anthropic suspended the highly capable Fable model, many users were left relying on a more verbose alternative. This repository provides a toolkit to steer Claude Opus toward the calmer, more direct working style of Fable. It includes a behavioral governor to suppress unnecessary self-auditing, a re-injection hook to keep your settings active, and a leak test to measure results from your own logs. It is a practical way to regain that efficient, result-first experience and get more work done faster.

🆔 @hackernewsgithubprojects

📦 hkuds/clawteam

Stop Managing Solo Agents: Unlock Swarm Intelligence with ClawTeam

Are you tired of juggling isolated AI agents that work alone? ClawTeam changes the game by transforming solo agents into a collaborative, self-organizing swarm. This powerful framework allows a leader agent to spawn specialized sub-agents that divide complex tasks, manage dependencies, and communicate in real-time. By utilizing git worktrees for workspace isolation and tmux for live monitoring, ClawTeam lets you automate everything from large-scale machine learning research to full-stack software engineering with a single command. It is time to step back, provide the vision, and let your agent team do the heavy lifting—ship your next project faster.

🆔 @hackernewsgithubprojects

📦 vasu-devs/justhireme

Take Control of Your Job Search with JustHireMe

Are you tired of drowning in endless, irrelevant job boards and black-box AI tools that just don't get you? Meet JustHireMe, a local-first desktop workbench designed to put you back in the driver's seat. It helps you scrape job leads from diverse sources, filters out low-quality postings, and ranks them based on your unique profile. By using local graph data and semantic matching, it generates tailored resume PDFs, cover letters, and outreach drafts right on your machine. Take command of your career with a private, explainable pipeline that turns the job hunt into a clear, manageable process today.

🆔 @hackernewsgithubprojects

📦 bisdom-cell/openclaw-model-bridge

Connect Any LLM to OpenClaw in One Command

Stop struggling with restrictive AI setups and take full control of your agentic workflow. This project is a powerful, production-tested middleware that lets you connect any major large language model to your OpenClaw environment instantly. Whether you prefer Qwen, OpenAI, Gemini, or others, this tool acts as a bridge that supports multimodal reasoning and seamless provider switching with zero third-party dependencies in its core runtime. It is designed to be highly reliable, featuring advanced tool governance, automated health checks, and a robust fallback system. Start building your custom AI agent and experience true, persistent automation today.

🆔 @hackernewsgithubprojects

📦 sunnweiwei/odyssim

Build Human Simulators with OdysSim

Have you ever wanted AI to truly simulate how humans think and behave rather than just providing simple answers? The OdysSim project is changing the game by building foundation models designed specifically for human behavior simulation. This repository offers a complete pipeline for midtraining and reinforcement learning, allowing you to train models that act with realistic social grounding. It features unique capabilities like learning from verbal feedback through LLM-judge critiques and multi-agent simulation loops. By bridging the gap between simple chatbots and authentic behavior, this framework helps researchers push AI closer to human-like interaction. Dive in and start simulating.

🆔 @hackernewsgithubprojects

📦 tolibear/goalbuddy

Stop Losing Your Place: Use GoalBuddy for Coding AI

Ever feel like your coding AI loses focus during long, complex tasks? Meet GoalBuddy, a powerful operating loop designed specifically for Codex and Claude Code. It transforms your workflow by giving every project a dedicated local workspace with a clear charter, live work boards, and a built-in verification loop. Instead of restarting your plan every time, GoalBuddy tracks your progress, handles subgoals, and keeps your work organized with receipts. It ensures your AI stays on target until the final goal is met. Install it today to keep your coding sessions focused, verified, and perfectly on track.

🆔 @hackernewsgithubprojects

📦 florianbruniaux/claude-code-ultimate-guide

Master Claude Code: The Ultimate Guide You Need

Are you tired of just copy-pasting AI configurations and hoping they work? Stop guessing and start mastering your development workflow. The Claude Code Ultimate Guide is a massive repository designed to take you from a total beginner to a power user through structured learning. With over two hundred quiz questions, production-ready templates, and deep dives into agentic workflows, this guide explains the reasoning behind every tool. You will learn how to design secure AI systems and master complex development methodologies. Level up your coding game today and gain the confidence to build your own custom agentic workflows.

🆔 @hackernewsgithubprojects

📦 zlab-princeton/llm-pruning-collection

Master LLM Pruning with This Powerful Research Collection

Ever wonder how to make large language models faster and more efficient without starting from scratch? The Princeton zLab has released a brilliant JAX-based repository that consolidates a wide array of state-of-the-art pruning techniques into one unified framework. By providing standardized implementations for methods like Minitron, Wanda, and SliceGPT, this collection solves the headache of inconsistent research environments. It includes everything you need for training on both GPUs and TPUs, plus evaluation scripts to measure performance accurately. Whether you are optimizing model size or speed, this is an essential toolkit for your compression experiments.

🆔 @hackernewsgithubprojects

📦 mig-nju/omnivideo-100k

Master Video Reasoning with OmniVideo-100K

Struggling to get AI to truly understand complex video content? Meet OmniVideo-100K, a powerful new dataset and automated synthesis engine designed to solve the common pitfalls of AI video analysis. By transforming raw footage into structured, entity-anchored scripts, this project helps models master intricate tasks like causal reasoning, future prediction, and deep context understanding. It is a game-changer for researchers aiming to bridge the gap between sight and sound with enhanced accuracy and coherence. Whether you are fine-tuning models or building new benchmarks, this framework provides the tools to boost performance significantly. Dive in and elevate your AI research today.

🆔 @hackernewsgithubprojects

📦 google-gemini/gemini-skills

Boost Your AI Coding Agents with Gemini Skills

Are you tired of your AI coding agents struggling with outdated knowledge? The Gemini Skills repository solves this by providing essential, up-to-date context specifically for building apps with the Gemini API. These skills act as a guide for your agents, ensuring they follow modern best practices for everything from real-time streaming with the Live API to using the latest interactions tools and managed agents. By integrating these skills, you help your AI generate more accurate, efficient code using current models and SDKs. Install these tools today to bridge the knowledge gap and power up your development workflow immediately.

🆔 @hackernewsgithubprojects

📦 master-cai/research-paper-writing-skills

Master Research Paper Writing with AI

Struggling to get your academic paper to flow? This GitHub repository offers a powerful, specialized skill package designed to help you draft and refine high-quality research papers in machine learning, computer vision, and natural language processing. By integrating these structured workflows into tools like Claude Code or Gemini, you gain access to expert-curated templates, section-specific writing guides, and rigorous self-review checklists. It helps you perfect your abstract, logical flow, and claim-evidence alignment from a reviewer's perspective. Download these tools today to transform your writing process and produce clearer, more impactful research papers that stand out to reviewers.

🆔 @hackernewsgithubprojects

📦 nikmcfly/mirofish-offline

Run AI Social Simulations Locally with MiroFish Offline

Ever wanted to predict public reaction to a policy or report without relying on expensive cloud services? Meet MiroFish Offline, a powerful multi-agent simulation engine that runs entirely on your own hardware. By replacing cloud dependencies with Neo4j and Ollama, this tool lets you upload documents to generate hundreds of unique AI agents that simulate complex social dynamics, sentiment shifts, and arguments in real time. It is perfect for testing PR crises or creative experiments, giving you full control and privacy. Download it today to start building your own private, swarm-intelligence social simulations locally on your machine.

🆔 @hackernewsgithubprojects

📦 zts212653/clowder-ai

Build AI Teams with Clowder AI

Stop managing isolated AI agents and start building coordinated AI teams. This repository offers a sophisticated framework designed to move beyond simple automation by providing hard rails, soft power, and a shared mission for your AI agents. It features a comprehensive suite of tools for collaborative thinking, enterprise workflows, and complex task orchestration. With capabilities like thread-based synchronization, memory navigation, and automated quality gates, it helps you manage agent lifecycles effectively. Whether you are debugging code or performing deep research, this project provides the infrastructure to align your AI swarm. Level up your agent coordination strategy today.

🆔 @hackernewsgithubprojects

📦 angristan/macthrottle

Is Your Mac Thermal Throttling? Here is How to Tell!

Ever wonder if your Mac is slowing down because it is getting too hot? MacThrottle is a lightweight menu bar application designed specifically to solve this mystery. It monitors your system thermal pressure in real time, alerting you the moment your performance begins to suffer from heat. Beyond simple alerts, it provides a detailed history graph of thermal states, CPU temperatures, and fan speeds. It reads data directly from your hardware, ensuring you get accurate insights into your Mac usage without needing any special permissions. Take control of your machine and stay informed with this simple, effective diagnostic tool.

🆔 @hackernewsgithubprojects

📦 nvk/llm-wiki

Build an AI-Maintained Knowledge Base with LLM Wiki

Tired of your AI assistant forgetting everything you teach it? Most AI sessions start from zero, forcing you to reconstruct knowledge every single time. The LLM Wiki repository solves this by turning your raw documents into a persistent, AI-maintained knowledge base. Using powerful tools like parallel multi-agent research and thesis-driven investigation, it automatically ingests your sources, compiles them into structured markdown, and keeps everything interlinked and current. It handles the tedious bookkeeping so you can focus on the insights. Whether for research or projects, this workflow helps your AI compound knowledge over time. Start building your own evolving second brain today.

🆔 @hackernewsgithubprojects

📦 tontinton/maki

Meet Maki: The Ultra-Efficient AI Coding Agent

Stop wasting precious tokens on bloated AI coding tools. Maki is an incredibly fast, memory-efficient AI coding agent built in Rust that prioritizes performance and transparency. Unlike other tools that hide how they process your data, Maki uses smart tree-sitter parsing to generate compact file skeletons, drastically reducing context window usage while keeping you fully in control. It features a responsive terminal user interface, built-in support for major AI providers, and a flexible Lua-based plugin system for deep customization. Whether you are debugging or building, Maki keeps your workflow lean and lightning-fast. Give your coding agent an upgrade today.

🆔 @hackernewsgithubprojects

📦 accius/openhamclock

OpenHamClock: A Modern Future for Amateur Radio Monitoring

Are you ready to keep your amateur radio station running smoothly beyond the scheduled shutdown of the original clock? OpenHamClock is a modern, community-driven, open-source replacement designed to preserve and evolve this essential tool. It offers a powerful, customizable dashboard for monitoring live DX clusters, satellite tracking, solar conditions, and even local weather data. With built-in features like rig control, APRS integration, and a flexible layout system, it creates a seamless experience for operators globally. Join a growing community of contributors dedicated to keeping our stations connected and well-informed. Check out the project today to help shape the future of ham radio.

🆔 @hackernewsgithubprojects

📦 harumiweb/exstruct

Turn Your Messy Excel Sheets into Structured JSON for LLMs

Are you struggling to feed complex Excel spreadsheets into your AI models? Meet ExStruct, the powerful Excel structured extraction engine designed to bridge the gap between messy workbooks and intelligent pipelines. It parses tables, charts, shapes, and formulas directly into clean, structured JSON format, making it easier than ever for LLMs and RAG systems to reason over your data. Whether you need precise extraction on Windows with full Excel integration or a best-effort approach on other platforms, ExStruct keeps your workflow running smoothly. Unlock your spreadsheet data today and simplify your automated agent tasks with this essential toolkit.

🆔 @hackernewsgithubprojects

📦 nabidalam/road-to-machine-learning

Master Machine Learning: A Complete Roadmap from Zero to Hero

Stop feeling overwhelmed and start mastering machine learning with this comprehensive step-by-step guide designed to take you from absolute beginner to production-ready expert. This repository offers a massive collection of twenty-six structured learning modules and twenty-three hands-on projects, covering everything from Python fundamentals and statistics to advanced topics like neural networks, generative AI, and MLOps. Whether you are building house price predictors or deploying large language models, this project provides the clear roadmap and practical resources you need to succeed. Dive into this organized collection today, follow your chosen career path, and start building your own professional portfolio.

🆔 @hackernewsgithubprojects

📦 logo-cuhksz/trl-bench

Stop Guessing Tabular AI Performance: Meet TRL-Bench

Ever wonder how different tabular AI models really compare? Evaluating these systems is often messy, because they are usually tied to specific, closed pipelines. That is where TRL-Bench comes in. It is a powerful, multi-granular benchmark that standardizes how we test tabular encoders. By forcing models to export standardized row, column, or table embeddings, it allows for direct, fair comparisons across twenty different models. Whether you are working on schema understanding or large-scale data enrichment, this project provides a consistent protocol for evaluating your models. It is the ultimate tool for transparency in tabular representation learning. Check it out to improve your workflow.

🆔 @hackernewsgithubprojects