GenieX is a production-grade, on-device generative AI inference SDK and CLI developed by Qualcomm for Snapdragon platforms. It provides a C core with a plugin architecture supporting two inference backends: llama.cpp (GGUF models on CPU/GPU/NPU) and QAIRT (pre-compiled bundles on Hexagon NPU). The system includes a Rust-based model manager that downloads, caches, and resolves models from Hugging Face, Qualcomm AI Hub, Docker Hub, and local filesystem, with resumable chunked downloads and progress reporting. It exposes multiple client interfaces: a Go CLI with subcommands for inference, an OpenAI-compatible HTTP server (streaming chat/completions, tool calls), Python bindings with high-level AutoModel factories, Android bindings via JNI/Kotlin, and Go bindings. The architecture is cleanly layered: C SDK → plugin backends → hardware, with a separate model manager process. The codebase demonstrates strong engineering practices: extensive test suites (unit, integration, device-level), CI/CD pipelines, versioned releases, and cross-platform support (Linux ARM64, Windows ARM64, Android). While primarily an application, it includes benchmark infrastructure and quality evaluation tools that could be used for research purposes.
Goeland is a self-hosted RSS/Atom feed to email digest tool written in Go. It fetches feeds from RSS/Atom/JSON sources, Imgur tags, and Miniflux instances, applies a composable chain of filters (e.g., unseen, today, digest, language, retrieve, replace), and sends beautiful responsive HTML emails via SMTP or writes to HTML files. It runs as a one-shot CLI command or as a persistent daemon with per-pipe cron scheduling. The architecture is modular: commands are defined with Cobra, configuration is read via Viper (TOML files with env overrides), logging uses Logrus with structured output support, and unseen entry tracking uses BoltDB. External dependencies include an SMTP server, Miniflux REST API, Imgur API, and optional CSS themes. The codebase demonstrates strong engineering practices: clean package boundaries, thorough unit tests (filter_test.go covers all filters including edge cases), security-conscious HTTP fetching with SSRF protection, internationalization support, and automated CI/CD with Docker multi-arch builds. It is production-stable and actively maintained.
oq is a terminal-based OpenAPI Specification viewer that allows users to visualize and navigate OpenAPI 3.0, 3.1, and 3.2 specifications in JSON or YAML format. The application reads OpenAPI specs from files or stdin, parses them using the libopenapi library, and presents the information in an interactive terminal UI with views for endpoints, components, and webhooks. Users can navigate using keyboard shortcuts, fold/unfold sections to see details, and cycle between different views. The architecture separates OpenAPI processing logic (in oas.go) from the terminal UI (built with bubbletea), demonstrating good separation of concerns.
This repository is a collection of LuCI applications (luci-app-*) for OpenWrt/iStoreOS, providing web-based management UIs for Docker services, system monitoring (netdata), and third-party tools (OpenClaw AI gateway, Sunpanel, mymind). It operates as an OpenWrt package feed: GitHub Actions workflows build IPK packages from the source tree, and a Go CLI tool (`forkapp`) clones and uploads LuCI apps to routers. The target platform is iStoreOS, with feeds configured from nas-packages and istore sources.