Start with the job, then find an open-source answer

Psysonic is a desktop music client for self-hosted libraries, designed primarily for Navidrome while also connecting to Gonic, Airsonic, LMS, and other Subsonic-compatible servers. It analyzes loudness, waveforms, and tempo on the user's machine to support normalization, playback-speed controls, and content-aware AutoDJ transitions. Navidrome users also get direct access to server-native features such as smart-playlist editing and playback reporting. Other servers may work well, but advanced behavior depends on the capabilities they expose.

chatgpt-imagegen lets a CLI or agent use a signed-in ChatGPT account to generate and edit images, apply shared styles, and turn model-produced sprite sheets into WebP or GIF animations. Its default backend drives a logged-in browser, with a Codex-based fallback available. The tool relies on undocumented internal endpoints, so it may break as the platform changes and should not power a public image-generation service.

Ascendara manages the discovery, download, organization, and launching of pre-installed games while reducing manual extraction and setup steps. Its Windows and Linux application offers two-click downloads for supported titles, fast local indexing, game information, browser integration, cloud backups, and encrypted remote progress viewing. The catalog is currently populated mainly from third-party sources such as STEAMRIP. Builds made from the public source also have limitations: games cannot be launched in development mode, and some reporting or analytics APIs are unavailable.

LLM-God places several AI web interfaces in one desktop window so the same text or screenshot can be sent to ChatGPT, Gemini, Claude, Grok, DeepSeek, and Copilot with a keyboard shortcut. Each embedded service still requires its own signed-in account and remains subject to the provider's normal usage limits. The supported release is currently for Windows, and the installer is not code-signed, so Windows may display an untrusted application warning.

Nucleus extends Compose Multiplatform with native windowing, operating-system integration, signing, and installer configuration through a Gradle DSL. A Kotlin codebase can be compiled and packaged for macOS, Windows, and Linux, running either on the JVM or as a GraalVM Native Image. It is aimed at teams that want a Kotlin-based desktop stack without giving up platform features. Projects require JDK 17 or later, Kotlin 2.0 or later, and Gradle 8.0 or later.

Joplin Note Tabs adds a tab panel to the Joplin desktop app so you can keep several notes open and pin important ones. Tabs can include navigation controls, checklist progress, and breadcrumbs, and they can be arranged horizontally or vertically and controlled with shortcuts. Pinned tabs persist in the local database but do not sync to other devices, and the plugin requires Joplin 1.8.2 or newer.

vault is a desktop organizer for links, Markdown notes, and original-quality images stored in ordinary local folders. Its browser extension sends saved content to the desktop app through a local API, where items can be grouped into folders or separate vaults, searched, renamed, pinned, imported, and exported. It works without an internet connection and does not include built-in synchronization, although users can sync the vault folder themselves with another storage service.

mlx-serve combines a native Zig inference server for Apple Silicon with the MLX Core menu-bar app. Users can download and manage MLX or GGUF models, chat from the terminal or desktop UI, use tool-calling and agent features, and connect existing clients through OpenAI-, Anthropic-, or Ollama-compatible APIs. It does not require Python or cloud inference, but it is specifically designed for Apple Silicon Macs and remains limited by the machine's available memory and compute.

Paste a Spotify playlist, album, or track link and Sunnify reads metadata from Spotify’s public pages, finds matching audio through yt-dlp, and transcodes it locally to MP3, M4A, FLAC, Opus, or WAV with titles, artists, albums, years, track numbers, and artwork embedded. Its graphical app runs on Windows, macOS, and Linux without a Spotify account or API key. The project is presented as an educational demonstration and should be used only for content the user owns or is allowed to download, for personal and non-commercial use.

Unreal Engine Skills provides 27 focused guides for UE5 C++ work across gameplay, rendering, networking, animation, AI, world streaming, UI, build tooling, and testing. A shared project-context skill captures modules, target platforms, and conventions, while the guidance has been checked against Unreal Engine source for valid signatures and class relationships. It is designed for game developers who want agents to follow real APIs and project-specific patterns.

illo is an Agent Skill for turning articles, blog ideas, or abstract concepts into original print-style editorial illustrations led by a recurring mascot. It includes character packs, ten bundled visual looks, palette controls, and a character builder, while keeping each image focused on one metaphor. Rendering can use an authenticated Codex CLI or Grok CLI subscription, or a selectable pay-per-image model through OpenRouter.

This set of Claude skills expands a simple video concept into a detailed Seedance 2.0 prompt with second-by-second timing, camera direction, lighting, sound layers, source-media references, and opening hooks. It covers formats such as SaaS launches, faceless channels, personal brands, and short-form social clips. The skills create prompts for Higgsfield rather than rendering the final video themselves.

Use tld to map and maintain complex software architecture through a C4-inspired visual workspace. Diagrams can be edited in the UI or managed from the CLI, YAML files, and an agent skill, with Mermaid import and export plus links back to editors and GitHub. The standalone binary includes the web app and can run offline with SQLite or be self-hosted; PostgreSQL deployments require pgvector.

This repository collects independent agent skills for practical workflows such as image generation, book illustration, development harnesses, social monitoring, video download and transcription, 3D generation, and content research. Each skill lives in its own directory, so users can adopt only the capabilities relevant to their work.

compose-skill equips Codex, Cursor, and Claude Code with on-demand guidance for the Compose development lifecycle. Its references cover architecture, state and coroutines, navigation, networking, persistence, performance, accessibility, testing, build configuration, and distribution across Android, iOS, desktop, and web targets. It is an agent skill rather than a runtime library added to the app.

Kiwix JS PWA lets users download ZIM archives and browse resources such as Wikipedia, Project Gutenberg, TED Talks, Wikivoyage, and Stack Exchange while offline. It can be installed as an offline-capable progressive web app, used through packaged desktop builds, or self-hosted from a container. The relevant ZIM archive must be obtained first, because the offline content is served from locally stored files.

AIPointer opens a small overlay beside the cursor when you hold a global hotkey, then sends the nearby screen region, your question, and optional clipboard text or selected files to a configured vision model. It is suited to translating visible text, explaining code, identifying charts or objects, summarizing documents, and suggesting replies, with optional voice input. It runs on macOS, Windows, and Linux but requires internet access, your own provider key, and screen-capture permissions.

Readwise Skills turns Readwise and Reader data into reusable agent workflows. An installed agent can triage inbox articles, catch up on feeds in batches, quiz you on recently read material, draft book reviews, or build personalized reading pages, with an optional reader persona guiding later outputs. The skills require either the Readwise MCP server or the Readwise CLI to be installed and connected first.