Recently updated open-source projects

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.

Equip coding agents with 37 plugins covering SAP BTP, CAP, Fiori and UI5, ABAP, HANA, Analytics Cloud, Datasphere, AI Core, and related workflows. Relevant guidance activates from project context. The skill content is portable, but Claude-specific agents, hooks, commands, and MCP configuration act only as guidance or manual validators in clients that do not support those features natively.

JS Paint recreates the tools and menus of classic Microsoft Paint in a web application, with touch support, themes, extra file types, and accessibility-oriented controls. Users can load images from URLs, upload to Imgur, and create experimental multi-user sessions. Those sessions are explicitly not private and may interrupt each other or lose work, so they are unsuitable for dependable collaborative editing.

Fusio is a self-hosted, open-source API management platform and gateway written in PHP. It can expose databases, microservices, and custom PHP or JavaScript actions as controlled API products, with a developer portal, SDK generation, usage monitoring, subscription plans, and quotas. Native MCP support can present selected backend actions as tools for AI agents. Docker Compose is the documented quick start, with manual installation also available.

Install the skill, authorize Baidu Netdisk, and ask a compatible agent to upload, download, search, transfer, or share files. Operations stay inside the app-specific storage directory, deletion is unavailable, and potentially destructive changes require confirmation. The project is still in beta and supports macOS, Linux, and Windows through WSL rather than native Windows.

This project turns existing ComfyUI workflow JSON files into callable agent skills. It imports and normalizes a workflow, maps only selected parameters through a schema, and lets an agent run, submit, inspect, or check dependencies through a structured CLI. A visual web interface supports configuration and testing, with routing to local or remote ComfyUI servers for OpenClaw, Hermes Agent, Codex, Claude Code, and similar tools.

Rai Pal manages universal mods that work across multiple games built on compatible engines. It scans libraries from providers such as Steam, GOG, and Epic, detects installed games and their engines, then helps install, launch, and update the correct mod version. Its UEVR workflow can locate Unreal Engine titles and start the mod for a selected game. Detection for owned but uninstalled games relies partly on heuristics, so results can be imperfect.

Marginalia keeps PDFs, notes, spreadsheets, logs, and archives in an ordinary folder tree while adding library-style metadata and retrieval indexes. When asked a question, its agent finds the relevant material and reads original file windows before composing a cited answer. It is available as a desktop app, CLI, reusable HTTP backend, and MCP server, with local storage by default; users still need to configure an LLM provider for generated responses.

TidGi is a cross-platform desktop app for notes, GTD, and personal knowledge management built around Node.js TiddlyWiki. Create a local workspace or clone an online wiki, keep ownership of the underlying data, and back it up through a local Git repository or GitHub; public content can also become a blog. It suits users who value customization and the TiddlyWiki plugin ecosystem, while macOS users may need to explicitly trust the unsigned app.

Pin multiple skill repositories as flake sources, recursively discover their SKILL.md directories, namespace duplicate IDs, and select exactly which skills enter a bundle. Home Manager integration can sync that bundle to global or project-local destinations for Codex, Claude, Cursor, OpenCode, and other agents. Targets are opt-in rather than automatically installed, making the project best suited to Nix users who want reproducible skill management.

Use shared skills and runtime-specific prompts to apply Spec Kit in Claude Code, Codex CLI, GitHub Copilot CLI, or Gemini CLI. The workflow moves from project principles and feature specifications through clarification, technical planning, consistency checks, ordered tasks, and implementation, with a baseline path for existing codebases. Spec Kit and the chosen agent CLI must be installed before these skills can run.

color-expert is an installable skill for Claude Code, Codex, Cursor, Copilot, OpenCode, and other agents. A concise SKILL.md provides frequently needed facts, corrections, and practical guidance, while an index routes the agent to 144 deeper references on color spaces, APCA and WCAG accessibility, palettes, pigment mixing, historical theory, tools, papers, and transcripts. Trigger evaluations, realistic task prompts, and maintenance guidance help keep the collection organized. It is intended for explanations, color reviews, and implementation advice.

Use this Codex skill after a single-slide reference image has established the visual direction. It crops complex illustrations, icons, or maps that are impractical to redraw, while recreating titles, body text, cards, arrows, labels, and conclusion bars with editable PowerPoint objects. The workflow works page by page from individual PNG references and explicitly avoids placing the whole screenshot as a flat slide background, preserving editability where it matters.

SuperSEO Skills lets an agent start with a URL, keyword, or topic, research the page, search results, and competitors, then produce SEO audits, content plans, drafts, and other growth work. It is built for live, one-off research workflows rather than historical Search Console analysis or ongoing site monitoring.

Keinsaas Navigator is an open-source AI workspace for individuals and teams. It combines multi-provider chat with custom agents, MCP tools, visual workflows, and optional capabilities such as image generation, code execution, and realtime voice interaction. The application can run locally, in Docker, or on Vercel. A model-provider API key is the minimum requirement, while features such as web search, file storage, or particular integrations need their own service configuration.

TaskingAI is a backend platform for developing and deploying LLM-based agents. It brings models from multiple providers, tools, RAG systems, assistants, and conversation history under one console and unified API. Developers can prototype configurations in the UI, then integrate them into multi-tenant products through REST endpoints or the Python SDK. The community edition can be self-hosted with Docker Compose and requires configuration for the selected model services.

This kit helps creators choose AI video models, translate prompts between model formats, and reuse tested workflows for storyboarding and debugging. It combines a large prompt library with Claude skills, methodology guides, and automated monitoring of official endpoints, making it useful for comparing or moving work among Seedance, Kling, Sora, Veo, and related systems.

IINA+ is a macOS companion application that opens content from several video and live-streaming services in IINA and displays danmaku where supported. The compatibility table includes services such as Bilibili, Douyu, Huya, Douyin, and NetEase CC, but live status, video decoding, and danmaku availability differ by site. Users can install a release build, while source builds require the latest Xcode and the project's dependency setup script.