llm
The AI Operator Living in Your Web Page. JavaScript in-page GUI agent. Control web interfaces with natural language.
🪄One line of code, turns your website into an AI-native app. Users give natural language commands, AI handles the rest.
macOS menu bar app that exposes Apple's on-device Foundation Models as an OpenAI-compatible local API. Zero cloud. Zero dependencies.
Related contents:
Threat Designer is a GenerativeAI application designed to automate and streamline the threat modeling process for secure system design.
Threat Designer is an AI-driven agent that automates and streamlines the threat modeling process for secure system design. Harnessing the power of large language models (LLMs), it analyzes system architectures, identifies potential security threats, and generates detailed threat models—empowering developers and security professionals to incorporate security from the earliest stages of development.
Open Source Data Labeling.
The most flexible data labeling platform to fine-tune LLMs, prepare training data, or evaluate AI systems. Label Studio is a multi-type data labeling and annotation tool with standardized output format. Label Studio is an open source data labeling tool. It lets you label data types like audio, text, images, videos, and time series with a simple and straightforward UI and export to various model formats. It can be used to prepare raw data or improve existing training data to get more accurate ML models.
Lightning-Fast RL for LLM Reasoning and Agents. Made Simple & Flexible.
AReaL is an open-source fully asynchronous reinforcement learning training system for large reasoning and agentic models, developed by members from Tsinghua IIIS and the AReaL Team at Ant Group. Built upon the open-source project ReaLHF, we are fully committed to open-source principles by providing the training details, data, and infrastructure required to reproduce our results, along with the models themselves. AReaL aims to help everyone build their own AI agents easily and affordably. Our team loves milk tea because it's delicious, customizable, and affordable—we hope you enjoy our project just as much as you'd enjoy real milk tea. Cheers!
The vector database that gets smarter the more you use it — and now ships as a cognitive container.
RuVector is a High Performance, Real-Time, Self-Learning, Vector Graph Neural Network, and Database built in Rust.
Most vector databases are static — they store embeddings and search them. That's it. RuVector is different: it learns from every query, runs LLMs locally, scales horizontally, boots as a Linux microservice from a single file, and costs nothing to operate.
spec-driven, production-ready code generation.
Related contents:
LLM Vulnerability Scanner.
Test large language models against 210+ adversarial attacks covering prompt injection, jailbreaks, encoding exploits, and data extraction.
Related contents:
Cross-linguistic entity matching across early modern texts.
A prototype for a digital history project aimed at creating a searchable concordance, knowledge graph, and resources for better understanding premodern science and natural knowledge.
Related contents:
Learn Languages While You Watch. No More Memorizing.
Turn YouTube, Netflix, and Bilibili into your personal language tutor. Just watch naturally, and let AI help you understand and learn new words instantly.
Related contents:
Write agentic workflows in natural language markdown, and run them in GitHub Actions.
Related contents:
Context window optimization for LLMs. Reliable LLM outputs start with clean context.
A reliability layer for LLM context. Deterministic deduplication that removes redundancy before it reaches your model. Less redundant data. Lower costs. Faster responses. More efficient & deterministic results.
Related contents:
LLM Service Fingerprinting Tool.
Simple LLM service identification - translate IP:Port to Ollama, vLLM, LiteLLM, or 30+ other AI services in seconds.
Julius is an LLM service fingerprinting tool for security professionals. It detects which AI server software is running on network endpoints during penetration tests, attack surface discovery, and security assessments.
Unlike model fingerprinting tools that identify which LLM generated text, Julius identifies the server infrastructure: Is that endpoint running Ollama? vLLM? LiteLLM? A Hugging Face deployment? Julius answers in seconds.
Related contents:
AI-powered subdomain enumeration tool with local LLM analysis via Ollama - 100% private, zero API costs.
God's Eye is a powerful, ultra-fast subdomain enumeration and reconnaissance tool written in Go. It combines multiple passive sources with active DNS brute-forcing and comprehensive security checks to provide a complete picture of a target's attack surface.
Related contents:
Kubernetes-native AI serving platform for scalable model serving.
Related contents:
Intercept LLM API traffic and visualize token usage in a real-time terminal dashboard. Track costs, debug prompts, and monitor context window usage across your AI development sessions.
tokentap tracks token usage for LLM CLI tools with a live terminal dashboard. See exactly how many tokens you're using in real-time.
Query your data in plain English.
Turn natural language questions into SQL queries with a small, local model that matches cloud LLM accuracy.
We fine-tuned a small language model to convert plain English questions into executable SQL queries. Because it's small, you can run it locally on your own machine, no API keys, no cloud dependencies, full privacy. Load your CSV files, ask questions, get answers.
Human-like Document AI
PageIndex is a vectorless, reasoning-based RAG engine that mirrors how humans read, delivering traceable, explainable, and context-aware retrieval, without vector databases or chunking.
AI Code Security Anti-Patterns distilled from 150+ sources to help LLMs generate safer code.
A comprehensive security reference distilled from 150+ sources to help LLMs generate safer code.
Related contents:
A native command-line interface for working with Apple Core ML models on macOS. Inspect, run inference, benchmark, and manage Core ML models without Xcode or Python.
Related contents:
An all-in-one enhancement suite for Google Gemini - timeline navigation, folder management, prompt library, and chat export in one powerful extension.
A deterministic, autonomous loop runner for the Codex CLI. It processes exactly one task per iteration from a JSON backlog, with fresh context each run, and keeps a JSONL audit log for traceability.
Related contents:
Autonomous task execution plugin for LLM CLI - refactored modular architecture.
llm loop is a powerful plugin for the LLM CLI tool that enables autonomous, goal-oriented task execution. Unlike traditional single-turn LLM interactions, llm loop allows the AI to work persistently towards a goal by making multiple tool calls, analyzing results, and iterating until the task is complete.
Related contents:
AI-Powered Reverse Engineering with Ghidra.
OGhidra bridges Large Language Models (LLMs) via Ollama with the Ghidra reverse engineering platform, enabling AI-driven binary analysis through natural language. Interact with Ghidra using conversational queries and automate complex reverse engineering workflows.
Related contents:
Agent harness framework for building, running, and verifying LLM workflows.
Gambit helps you build reliable LLM workflows by composing small, typed “decks” with clear inputs/outputs and guardrails. Run decks locally, stream traces, and debug with a built-in UI.
The Best Agent Harness. Meet Sisyphus: The Batteries-Included Agent that codes like you.
Open Responses is an open-source specification and ecosystem for building multi-provider, interoperable LLM interfaces based on the OpenAI Responses API. It defines a shared schema, and tooling layer that enable a unified experience for calling language models, streaming results, and composing agentic workflows—independent of provider.
A simple, open format for giving agents new capabilities and expertise. Agent Skills are folders of instructions, scripts, and resources that agents can discover and use to do things more accurately and efficiently.
Related contents:
LLM-powered framework for deep document understanding, semantic retrieval, and context-aware answers using RAG paradigm.
It adopts a modular architecture that combines multimodal preprocessing, semantic vector indexing, intelligent retrieval, and large language model inference. At its core, WeKnora follows the RAG (Retrieval-Augmented Generation) paradigm, enabling high-quality, context-aware answers by combining relevant document chunks with model reasoning.
Semantic search for agents.
A calm, CLI-native way to semantically grep everything, like code, images, pdfs and more.
Related contents:
Understand Privacy & Legal Terms. Plain-language analysis of Terms & Conditions. Know what you're agreeing to before you sign up.
Related contents:
A dead-simple unix tool for lightweight open-source local agents.
Orla is a unix tool for running lightweight open-source agents. It is easy to add to a script, use with pipes, or build things on top of.
Related contents:
A curated catalogue of agentic AI patterns — real‑world tricks, workflows, and mini‑architectures that help autonomous or semi‑autonomous AI agents get useful work done in production.
Related contents:
Multi-agent orchestrator for Claude Code. Track work with convoys; sling to agents.
Related contents:
- Welcome to Gas Town @ Steve Yegge's Medium.
- Gas Town Decoded @ Andrew Lilley Brinker.
- How to think about Gas Town @ Steve Klabnik.
- Agent Psychosis: Are We Going Insane? @ Armin Ronacher's Thoughts and Writings.
- Gas Town’s Agent Patterns, Design Bottlenecks, and Vibecoding at Scale @ Maggie Appleton.
- Move Over Gas Town, Claude Has First-Party Agent Orchestration @ Andrew Lilley Brinker.
The World's First LLM-Native Language. the first intermediate language designed for machine authorship.
Reduce logs to their semantic anomalies.
Cordon uses transformer embeddings and density scoring to identify semantically unusual patterns in large log files, reducing massive logs to the most anomalous sections for analysis. Repetitive patterns (even errors) are considered "normal background." Cordon surfaces unusual, rare, or clustered events that stand out semantically from the bulk of the logs.
Related contents:
Layrr is the visual editor for real code. Design visually, edit any stack, own everything. A browser coding agent interface for selecting elements and sending instructions directly to Claude Code.
Open Source Locally Hosted Lovable with Full Stack Support. The Open-Source AI Development Platform Built for Self-Hosting.
AI-powered development environment with advanced agent orchestration - designed for complete data sovereignty and infrastructure control.
Unterminated Block Parsing.
Remend is a lightweight, standalone preprocessor that completes incomplete Markdown syntax.
Related contents:
Production-ready RAG in your infrastructure.
Skald gives you a production-ready RAG in minutes through a plug-and-play API, and then let's you configure your RAG engine exactly to your needs.
Our solid defaults will work for most use cases, but you can tune every part of your RAG to better suit your needs. That means configurable vector search params, reranking, models, query rewriting, chunking (soon), and more.
Related contents:
Metis is an open-source, AI-driven tool for deep security code review.
Metis is an open-source, AI-driven tool for deep security code review, created by Arm's Product Security Team. It helps engineers detect subtle vulnerabilities, improve secure coding practices, and reduce review fatigue. This is especially valuable in large, complex, or legacy codebases where traditional tooling often falls short.
Executable GenAI Prompt Templates.
Dotprompt is an executable prompt template file format for Generative AI. It is designed to be agnostic to programming language and model provider to allow for maximum flexibility in usage. Dotprompt extends the popular Handlebars templating language with GenAI-specific features.
Enforce the habit of self-documenting code through better commit messages.
smartcommit is an intelligent, AI-powered CLI tool that helps you write semantic, Conventional Commits messages effortlessly. It analyzes your staged changes, asks clarifying questions to understand the "why" behind your code, and generates a structured commit message for you.
The toolkit for AI devtools context engineering. Build with codebase mapping, symbol extraction, and many kinds of code search.
kit is a production-ready toolkit for codebase mapping, symbol extraction, code search, and building LLM-powered developer tools, agents, and workflows.
kit shines for getting precise, accurate, and relevant context to LLMs. Use kit to build code reviewers, code generators and graphs, even full-fledged coding assistants: all enriched with the right code context.
AI Browser Automation. Automate browser based workflows with AI.
Skyvern automates browser-based workflows using LLMs and computer vision. It provides a simple API endpoint to fully automate manual workflows on a large number of websites, replacing brittle or unreliable automation solutions.
Emdash is an orchestration layer for running multiple coding agents in parallel in isolated Git worktrees.
An orchestration layer for running multiple coding agents in parallel, each isolated in its own Git worktree. Run several agent instances concurrently to tackle independent subtasks or experiments.
Emdash lets you develop and test multiple features with multiple agents in parallel. It’s provider-agnostic (we support 10+ CLIs, such as Claude Code and Codex) and runs each agent in its own Git worktree to keep changes clean; when the environment matters, you can run a PR in its own Docker container. Hand off Linear, GitHub, or Jira tickets to an agent, review diffs side-by-side, and keep everything local.
Tokenflood is a load testing framework for simulating arbitary loads on instruction-tuned LLMs.
Tokenflood is a load testing tool for instruction-tuned LLMs that allows you to run arbitrary load profiles without needing specific prompt and response data. Define desired prompt lengths, prefix lengths, output lengths, and request rates, and tokenflood simulates this workload for you.
🔂 Run Claude Code in a continuous loop, autonomously creating PRs, waiting for checks, and merging
Automated workflow that orchestrates Claude Code in a continuous loop, autonomously creating PRs, waiting for checks, and merging - so multi-step projects complete while you sleep.
Fully automatic censorship removal for language models.
Heretic is a tool that removes censorship (aka "safety alignment") from transformer-based language models without expensive post-training. It combines an advanced implementation of directional ablation, also known as "abliteration" (Arditi et al. 2024), with a TPE-based parameter optimizer powered by Optuna.
Enchanted is iOS and macOS app for chatting with private self hosted language models such as Llama2, Mistral or Vicuna using Ollama.
Enchanted is open source, Ollama compatible, elegant macOS/iOS/visionOS app for working with privately hosted models such as Llama 2, Mistral, Vicuna, Starling and more. It's essentially ChatGPT app UI that connects to your private models. The goal of Enchanted is to deliver a product allowing unfiltered, secure, private and multimodal experience across all of your devices in iOS ecosystem (macOS, iOS, Watch, Vision Pro).
Related contents:
Chat for Ollama, Private AI Chat.
Empowering LLM researchers and hobbyists with seamless control over self-hosted models. Connect remotely, customize prompts, manage chats, and fine-tune configurations. All in one intuitive app.
Related contents:
A desktop app to easily run Large Language Models locally.
CLI tool – estimates LLM tokens/costs and runs provider-aware load tests for OpenAI, Anthropic, OpenRouter, or custom endpoints.
A fast, CLI-based tool to estimate token usage and API cost for prompts targeting various LLM providers (OpenAI, Claude, Mistral, etc.). Built in Rust for performance, portability, and safety.
AI-powered git commit message rewriter using GPT.
Automatically rewrite your entire git commit history with better, conventional commit messages using AI. Perfect for cleaning up messy commit histories before open-sourcing projects or improving repository maintainability.
Ollama Manager App.
Unofficial Ollama manager app for macOS, iOS, iPadOS, and visionOS, featuring server management, model management, and simple chat feature.
Related contents:
Fast, stateless LLM-powered assistant for your shell: qq answers; qa runs commands.
qq means quick question. qa means quick agent. Both are easy to type rapidly on QWERTY keyboards with minimal finger movement. That makes interacting with LLMs faster and more natural during real work.
Web Codegen Scorer is a tool for evaluating the quality of web code generated by LLMs.
JSON for LLM prompts at half the tokens. Spec, benchmarks & TypeScript implementation.
Token-Oriented Object Notation is a compact, human-readable serialization format designed for passing structured data to Large Language Models with significantly reduced token usage. It's intended for LLM input, not output.
TOON's sweet spot is uniform arrays of objects – multiple fields per row, same structure across items. It borrows YAML's indentation-based structure for nested objects and CSV's tabular format for uniform data rows, then optimizes both for token efficiency in LLM contexts. For deeply nested or non-uniform data, JSON may be more efficient.
Related contents:
Large language model made in Europe built to support all official 24 EU languages.
The EuroLLM project includes Instituto Superior Técnico, the University of Edinburgh, Instituto de Telecomunicações, Université Paris-Saclay, Unbabel, Sorbonne University, Naver Labs, and the University of Amsterdam. Together they created EuroLLM-9B, a multilingual AI model supporting all 24 official EU languages. Developed with support from Horizon Europe, the European Research Council, and EuroHPC, this open-source LLM aims to enhance Europe’s digital sovereignty and foster AI innovation. Trained on the MareNostrum 5 supercomputer, EuroLLM outperforms similar-sized models. It is fully open source and available via Hugging Face.
Related contents: