Explore the directory

A lossless GPU-memory compression codec for AI model weights and KV caches, decoding compressed bf16 numbers on the GPU during matrix multiplication and attention so models fit in less memory.

• Inference

A Claude Code skill that adds multimodal vision to text-only model backends, letting the agent analyze pasted screenshots, UI states and charts through a configurable external vision model.

• Agent skills

A template-first CLI toolkit for local AI agent application development, generating an editable starter project from a locked template catalog and giving it a consistent lifecycle of commands.

• Agent frameworks

A Go framework for building multi-agent AI systems, with a streaming-first API, built-in memory and retrieval-augmented generation, and sequential, parallel, DAG, and loop orchestration patterns.

• Agent frameworks

A self-hosted AI workflow automation platform that reviews pull requests, generates tests, triages issues, and keeps documentation and translations in sync, across GitHub, GitLab, Gitea, and Bitbucket.

• Git
• CI/CD

A VS Code extension and local proxy that switches the AI model behind GitHub Copilot Chat instantly from a web dashboard, without editing configuration files or reloading the editor.

• LLM gateways
• Editors

An MCP server exposing the full Ollama SDK as tools, so Claude Desktop, Cline, and other MCP clients can drive local or Ollama Cloud language models directly.

• MCP servers

A static, zero-backend web site giving Chinese-language, six-dimension summaries of all 5,352 accepted ICLR 2026 papers, browsable by a two-level topic hierarchy, with the full generation pipeline included.

• Papers

A Python toolkit giving a single OpenAI-compatible interface across Chinese large language models, so frameworks such as LangChain and LlamaIndex can call them without model-specific adapter code.

• LLM gateways

A Unix command-line wrapper over Jina AI's search, read, embedding, and reranking APIs, designed so an AI agent with shell access can call one consistent tool instead of many separate API integrations.

• RAG
• CLI tools

An example retrieval-augmented-generation project demonstrating question answering over an internal document set using a locally deployed large language model.

• RAG
• Tutorials

A CI-native security scanner for MCP servers that simulates attacks, detects schema drift, and produces a health score and compliance evidence before other agents come to depend on the server.

• MCP servers
• Security

A lightweight Python AI agent framework built on event-driven, graph-based orchestration, with multi-channel messaging, cron scheduling, and streaming output.

• Agent frameworks

A prototype catalog of agent-callable tools for the apps Zapier integrates with, where each connector is simultaneously an installable agent skill and a native MCP tool, usable with or without Zapier.

• MCP servers
• Agent skills

An open-source toolkit for building RAG pipelines that converts PDFs to Markdown, cleans the result, and lets a user visually inspect and compare chunking strategies before indexing.

• RAG

A community-maintained hardware plugin that lets the vLLM inference engine run on MetaX GPUs, using MetaX's CUDA-like MACA backend for near-native CUDA performance.

• Inference

A cognitive memory database for AI agents that actively manages what it stores, consolidating duplicates, detecting contradictions, and fading out stale memories through temporal decay.

• Agent memory
• Databases

A command-line tool implementing Retrieval-Augmented Fine-Tuning, fetching a person's public writing from sources such as Substack, X, or RSS to build a combined fine-tuning and retrieval dataset.

• RAG
• Models

A terminal UI for the Beads issue tracker, with its own boolean query language for searching issues, browsing dependency trees, and building custom kanban views without leaving the terminal.

• Terminals

A curated set of 88 Microsoft security skills for AI coding assistants, covering Defender, Sentinel, Entra, and Purview, so an agent gives grounded, product-specific guidance instead of generic advice.

• Agent skills
• Security

A self-hosted personal AI agent with persistent identity, self-authored goals, scheduled autonomous activity, and a private journal, reachable over Telegram, Discord, Slack, Feishu, iMessage, and webhooks.

• Agent frameworks

A large collection of skills, agents, and enforced quality hooks for AI coding assistants, installing machine-checked safety gates such as credential scanning and required test runs across many platforms.

• Agent skills

A comprehensive, continually updated collection of process reward models, models that score each reasoning step rather than only the final answer, for mathematical and other LLM tasks.

• Curated lists
• Papers

A multi-language runtime that adds visibility and policy control over an agent's tool and LLM calls without modifying the existing agent stack, exporting events and traces to observability tools.

• AI evals