Trace, evaluate, monitor, and manage LLM, agent, and ML workflows.
Project overview
MLflow records and manages the development and operation of LLM applications, AI agents, and machine-learning models. Teams can start a server, enable automatic tracing, inspect calls and metrics in the UI, and use evaluation, prompt optimization, experiment tracking, model registry, and deployment workflows to improve their systems. It can run locally, on private infrastructure, or in cloud environments, making it useful when generative AI and traditional model lifecycles need a shared operational layer.
Automate and test Chromium, Firefox, and WebKit through one API.
Project overview
Playwright drives Chromium, Firefox, and WebKit through one API for end-to-end testing and general browser automation. It can be used as a test runner or library, through a CLI designed for coding agents, or through an MCP server for LLM-driven workflows. Its VS Code extension also supports test recording, debugging, locator picking, and trace inspection.
Build and ship compact cross-platform desktop apps with TypeScript.
Project overview
Electrobun gives TypeScript developers one toolkit for creating, packaging, updating, and distributing desktop applications. Start from an npm template, use Bun for the main process and webview bundle, and produce compact self-extracting releases with small binary-diff updates. Official platform support covers macOS 14 and later, Windows 11 and later, and Ubuntu 22.04 and later.
Research an industry and generate an installable domain-knowledge Agent Skill.
Project overview
master-skill researches a chosen niche across six tracks: leading practitioners, tool maps, workflows, canonical knowledge, information sources, and terminology standards. It then distills the material into a runnable Agent Skill with companion shell commands, source checks, claim tracing, quality gates, and installation tooling. Generated skills can be installed for Claude Code, OpenClaw, Codex, or Hermes. The repository describes a full research and generation run as taking roughly 30 to 60 minutes.
Build native cross-platform desktop apps with Kotlin and Compose.
Project overview
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.
Add progressively loaded Agent Skills to Pydantic AI applications.
Project overview
Attach skill directories to an agent and let it discover metadata first, then load full instructions, resources, or scripts only when needed. The package supports standard filesystem skills as well as skills created programmatically in Python, exposed through capability or toolset integrations. It also validates skill structure, accepts multiple source directories, and includes safeguards against path traversal.
Build and package web, desktop, and mobile apps using only Python.
Project overview
Flet lets Python developers create iOS, Android, Windows, Linux, macOS, and web applications from one codebase without first learning Dart, Swift, Kotlin, HTML, or JavaScript. Flutter renders the interface with a large set of Material and Cupertino controls. Web apps can run client-side through WebAssembly and Pyodide or as server-driven Python applications with real-time updates, and built-in tooling packages projects for each target platform.
Create skill evaluations, run benchmarks, and compare performance across models.
Project overview
Waza is a Go CLI for scaffolding Agent Skills evaluation suites, running benchmarks, and comparing outcomes across models or executors. Its readiness checks cover frontmatter compliance, token budgets, evaluation files, and agentskills.io specification rules, and the workflow can be integrated into CI. It is aimed at skill authors and teams that need repeatable quality measurements and regression checks.
Build synchronized AI agents visually or through a TypeScript SDK.
Project overview
Inkeep Agents lets non-technical teams design agents on a drag-and-drop canvas while engineers define the same agents in a TypeScript SDK, with two-way synchronization between both formats. It supports customer assistants, internal copilots, and multi-agent workflow automation, along with MCP credential handling, embeddable chat components, tracing, and Vercel or Docker deployment. The platform can be self-hosted, but it uses Elastic License 2.0 with supplemental terms rather than a permissive open-source license.
Control, automate, and operate Android devices or fleets from one integrated platform.
Project overview
FIRERPA combines Android remote desktop, UI automation, OCR and image matching, traffic capture, Frida, proxy networking, and device operations behind one on-device service and Python client. It can work with or without root and expose controls through a browser, APIs, MCP clients, or a natural-language agent. The platform targets automation engineering, mobile security work, and long-running device farms, so its breadth and deployment complexity go well beyond a simple phone mirroring utility.
Build extensible browser and desktop IDEs for multiple languages.
Project overview
Eclipse Theia is a framework for teams building their own developer tools or cloud IDEs, rather than a single fixed editor. It can be composed into browser-based or desktop applications, offers a flexible extension architecture, and supports the VS Code extension protocol. The project also documents application composition, Theia extensions, VS Code-style extensions, and desktop packaging.
Generate consulting-style PowerPoint decks with structured layouts and automated quality gates.
Project overview
MCK PPT Design Skill gives AI agents a consulting-style presentation system backed by a Python generation engine. Its workflow locks each slide's layout and takeaway in an outline, loads only the relevant design guidance, calls a library of slide and chart patterns, and runs machine-readable checks before and after rendering. It suits repeatable business-deck production and requires Python with python-pptx and supporting dependencies.
Ask questions across long videos with multimodal retrieval and scene-level context.
Project overview
VideoRAG is an open-source framework for retrieving and answering questions over long-form video, with the Vimo desktop interface for importing one or more videos and chatting about their contents. It also includes the LongerVideos benchmark and evaluation code for researchers working on long-context video understanding. Prebuilt Vimo downloads are still marked as coming soon, so current users generally need to start the Python backend and Electron frontend from source.
An open-source framework that evolves coding-agent skills from failed benchmark trajectories.
Project overview
EvoSkill is an open-source framework for discovering and evolving reusable coding-agent skills. It runs agents on benchmark tasks, analyzes failed trajectories, proposes skill and system-prompt variants, and evaluates them on held-out data. Projects can use Claude Code, Codex, OpenCode, OpenHands, or Goose with CSV question-and-answer sets or containerized Harbor tasks. It suits teams with a meaningful evaluation set and requires the selected agent runtime, model-provider credentials, and supporting environment.
Turn agent-run learnings into readable, editable, reusable Skill files.
Project overview
Acontext is an open-source memory layer that captures reusable lessons from AI-agent runs and stores them as ordinary Markdown Skill files. Developers can read, edit, version, mount, or share those memories across models and frameworks without relying on embeddings or a closed memory format; agents retrieve the needed files through tools and reasoning. Users can connect to the hosted service with an API key or self-host the backend with Docker, which requires an OpenAI API key.
Add local visual analysis, recognition, and smart alerts to camera feeds.
Project overview
DeepCamera is an open-source platform for running pluggable AI camera skills locally, including scene analysis, object detection, person re-identification, and fall detection. Skills use a common protocol and can target different model and hardware backends across Apple Silicon, NVIDIA, Intel, AMD, and CPU systems. Its Aegis desktop companion connects camera sources, installs skills, and delivers alerts or camera conversations through Telegram, Discord, or Slack. The catalog is active and still expanding, so capability maturity may vary by skill.
Structure AI-assisted engineering from requirements through implementation, testing, and review.
Project overview
Agentic Engineering Framework packages requirements clarification, system design, code generation, testing, and review as reusable commands, skills, and specialist agents. Teams can add project knowledge and coding standards, while smaller fixes can invoke only the troubleshooting, implementation, or performance workflow they need. The files are installed into supported coding-agent directories such as Claude Code or Codex CLI.
Turn an idea into a tested, packaged AI skill for several agent platforms.
Project overview
Skill Creator Ultra turns a workflow idea into a complete AI skill through an interview or a one-shot quick mode. Its eight-stage process extracts the knowledge, chooses an architecture, generates the package, runs tests and multidimensional evaluation, scans for security issues, iterates, and prepares distribution formats for several agent platforms. It is designed for users who do not already know YAML or SKILL.md, with English and Vietnamese as the documented interaction languages.