Explore the directory

A plug-and-play token-compression framework for video LLMs that dynamically allocates a frame token budget, cutting a model such as LLaVA-OV's generation time by over 70 percent while keeping most accuracy, unifying several published baselines.

• Inference
• Papers

A block-style, Scratch-inspired programming environment for building multi-agent simulations, built on Pharo Smalltalk, where agents such as turtles or termites follow simple rules that produce emergent behaviour, in the StarLogo tradition.

• Languages

A local, git-versioned memory system for AI coding agents that captures conversations, compiles them into durable project knowledge, and recalls context through a local MCP server, stored as plain Markdown with on-device embeddings, no database.

• Agent memory

A coding agent that runs entirely on an Apple Silicon Mac through MLX, using one 27-billion-parameter local model with no API key, an agent loop that owns the inference backend process directly for a stable KV cache, and a terminal UI.

• Coding agents

A curated, continuously updated list of papers applying large language models to prognostics and health management, covering fault diagnosis and remaining-useful-life prediction, each entry giving the English and Chinese titles, year, venue and code link where available.

• Papers

An enterprise AI agent platform on LangGraph and LangChain separating a governance control plane, covering auth, project management and audit, from an agent runtime handling graph orchestration, model assembly, tools, MCP and skills.

• Agent frameworks

A diffusion language model that decodes many tokens in parallel by revising its own earlier predictions, aiming to preserve output quality at high decoding parallelism.

• Models
• Inference

A Windows tool for red-team engagements that searches a host's filesystem, services and scheduled tasks for .NET assemblies that could serve as payloads or be used in post-exploitation.

• Security

A set of runtime patches applied to vLLM in memory, without forking or rebuilding it, that speed up inference for the Qwen3.6 and Gemma 4 model families on consumer NVIDIA GPUs with 24GB or more of memory.

• Inference

A reference architecture for running a council of AI coding agents such as Claude, OpenCode and Crush across one codebase with work-queue delegation, automated PR review and hardened CI, alongside research packages for sleeper-agent detection.

• Agent frameworks
• Security

A compliance and security scanning tool for AI applications, built around Chinese regulations, scanning a project in one command for foreign model endpoints, hardcoded keys and exposed personal data, with runtime interception of prompt injection.

• Security

A set of 20 MCP tools that index a whole Dynamics 365 Finance and Operations codebase of X++ classes, tables, forms and labels, so Copilot or Claude Code can check a method signature or form pattern against real metadata before generating code.

• MCP servers

A weight-streaming layer for Apple's MLX framework that runs language models larger than a Mac's RAM by loading weights from SSD as each layer executes, without quantizing them.

• Inference

A local-first AI coding agent living in a VS Code or Cursor sidebar that sends every request to the user's own machines, running against Ollama, LM Studio or a self-hosted OpenAI-compatible endpoint, with no account, cloud connection or per-token cost.

• Coding agents

A TypeScript and JavaScript library of small, composable building blocks for LLM-powered functions, including prompt templates, output parsers, conversation state and provider-agnostic calls across OpenAI, Anthropic, xAI, Google, AWS Bedrock, Ollama and DeepSeek.

• Agent frameworks

An MCP server that gives Claude Code an entire malware-analysis lab of over 300 tools, covering decompilation, symbolic execution, interactive debugging, YARA and capa signatures, emulation and .NET, Go and Rust analysis, driven from natural-language prompts.

• MCP servers
• Security

An agentic development workflow that pairs Claude Code as the executor with Codex as an adversarial reviewer, taking a requirement through research, planning, development and cross-model review with mechanical compile-and-test gates before each atomic commit.

• Coding agents

A research framework for repository-level code generation that retrieves context from external libraries as well as the local repository, using a knowledge graph over both.

• RAG
• Papers

A teaching project that builds a WeChat mini-program from scratch with Taro, React, FastAPI, LangChain, LangGraph and DeepSeek, where a single sentence produces AI-generated learning quizzes, with a retrieval-augmented mode for private documents and a full video and text course.

• Tutorials

A memory system for AI agents that stores facts, episodes and raw sentences in a three-layer graph backed by a single SQLite database by default, so agents can recall not just what a user said but how a preference or pattern changed over several sessions.

• Agent memory

A lightweight native C++ inference runtime for NVIDIA's Nemotron speech model family, covering streaming speech recognition, speaker diarization, speech translation and text-to-speech, powered by ggml with broad hardware support.

• Inference
• Speech

A public, symptom-first registry of LLM serving-path bugs that silently produce wrong benchmark numbers, covering chat templates, tool parsers, quantization kernels and eval harnesses, each entry with a reproduction status and the check that catches it.

• Inference
• AI evals

A terminal-based AI coding agent platform, distributed as a small native installer, that talks directly to a chosen provider or a local Ollama model, asks approval before writing files or running commands, and bundles a terminal, editor and file explorer.

• Coding agents

A curated collection of AI agent frameworks, developer tools and automation resources compiled by AI Pulse Georgia, searchable in natural language from Claude Code, Cursor or any MCP-compatible client through a bundled MCP server, or from the terminal with an accompanying CLI.

• MCP servers
• Curated lists