Claude Code is Anthropic's agentic coding tool for the terminal, IDEs, and GitHub. It uses natural-language commands to understand a codebase, create and read files, execute commands, run tests, explain code, manage Git workflows, and handle routine development tasks. It can also load persistent project context, run custom slash commands, use plugins with custom commands and agents, and operate with configurable autonomy while leaving actions such as final pull-request merging to a human. The official repository documents installation for macOS, Linux, and Windows, and identifies npm installation as deprecated.
Cloudflare R2 is Cloudflare's object and file storage service, providing an S3-compatible API for storing and serving large objects and uploaded files. It is a low-latency storage option integrated with Cloudflare's network and developer platform and is used by self-hosted screenshot-sharing software to store uploads.
Doop is an open-source multiplayer design canvas where people and AI agents create and edit designs together. Each canvas contains frames that render real HTML in sandboxed iframes; humans edit in the browser, while agents connect through the built-in Model Context Protocol (MCP) server and can create frames, stream HTML in chunks, inspect screenshots, and revise designs. The app synchronizes cursors, presence, frame edits, agent status, comments, tasks, and activity over WebSocket rooms, and includes a server-side Doop Agent that can process queued cards, mentions, and feedback through specialist roles. Canvases are private by default, with email invitations or optional link sharing, and agents inherit the access of the user who authenticates them. Doop also provides design-memory features for exemplar frames, decisions, and proposed style rules, and can be self-hosted with Docker Compose or Bun using embedded Postgres through PGlite.
ego lite is a macOS browser from Citro Labs designed for AI-agent browser automation alongside a user's normal browsing. It gives each agent or task an isolated Space within the same browser, allowing multiple tasks to run in parallel without taking over the user's tabs; users can observe, take over, or stop an agent's Space. Through the ego-browser skill, agents such as Claude Code, Codex, Cursor, or custom agents can call in-page JavaScript tools including snapshot, fill, click, wait, navigate, and capture; an agent can compose several operations into one JavaScript execution rather than repeatedly issuing CLI commands. During setup, optional Chrome-data migration can transfer existing logins, cookies, extensions, and bookmarks, while browsing data remains on the user's device. The browser is distributed as a separate free download, the repository is MIT-licensed, and Windows and Linux support are listed as planned.
Electron is an open-source framework for building cross-platform desktop applications with JavaScript, HTML, and CSS. It is built on Node.js and Chromium, provides binaries for macOS, Windows, and Linux, and is used by applications such as Visual Studio Code. The project is MIT-licensed, maintained under the OpenJS Foundation, and distributed via npm.
Git is an open-source, fast, scalable, distributed revision control system. It stores project changes as versioned commits and provides repository history, diffs, branching, isolated work trees, and rollback capabilities for reviewing or reverting changes. The project provides both high-level version-control operations and access to its internal mechanisms. It is covered primarily by the GNU General Public License, with some components under compatible licenses, and was originally written by Linus Torvalds with contributions from a wider developer community.
Google 3D Tiles is a Google service that provides 3D geographic tiles covering the globe.
hayamimi is a local, real-time multilingual speech-to-text tool that produces live subtitles, a browser dashboard, speaker labels, and optional translation without a GPU or cloud API. It detects the language of each utterance and routes it to a dedicated specialist model for Japanese, Chinese, Korean, Cantonese, English, and European languages, with a broad Omnilingual ASR fallback for other languages. The models run as quantized INT8 ONNX models through sherpa-onnx, without PyTorch or CUDA. The pipeline updates partial subtitles while speech is in progress, emits finalized lines shortly after speech stops, and can perform a second batch-decoding pass after silence to produce a refined transcript. Speaker labeling uses CAM++ embeddings and is intended for turn-taking rather than full diarization. The `--serve` option starts a local HTTP server with a live dashboard and an OBS browser-source overlay. Audio can also be supplied over WebSocket, and model eviction keeps resident models below a configurable memory cap, with the default under 2 GB of RAM. The project reports 5.8% character error rate on its Japanese broadcast-audio scorecard and 10–50× real-time processing on a six-core desktop CPU.
Kaneo is an open-source, self-hosted project-management tool for teams, focused on issue management and kanban-style workflows. It is designed as a simpler alternative to more complex project-management platforms, keeping project data on the user's infrastructure. The project can be deployed with the drim command-line tool, Docker Compose, or a Kubernetes Helm chart. Its Docker Compose setup runs Kaneo with PostgreSQL, and the project is distributed under the permissive MIT license.
Kubernetes (K8s) is an open-source container orchestration system hosted by the Cloud Native Computing Foundation. It manages containerized applications across multiple hosts, providing mechanisms for deploying, maintaining, scaling, and scheduling workloads. Applications can be configured with YAML and deployed across hybrid-cloud environments; the platform is also used for production operations involving storage, databases, networking, and security hardening.
LatticeDB is an embedded, single-file property-graph database written in Zig for local applications. It combines relationship traversal, HNSW vector similarity search, and BM25 full-text search in one query language and engine, allowing graph, semantic, and textual queries over the same dataset. It also provides durable named event streams and a graph changefeed through the same transaction and write-ahead-log path as graph writes. The database operates without a server or configuration and is designed for one owning process on one machine with WAL-backed durability. It is distributed through a CLI and bindings or packages for Python, TypeScript/Node.js, Java, and Go. Graph RAG, agent memory, and local knowledge tools are documented as example workloads rather than the engine's definition.
A personal AI learning system distributed as a pi configuration. It encodes a teaching philosophy and learning process in skills, including teaching and diagram-based visualization, and adds extensions for question popups, graded quizzes, Markdown session logs, and visualization tools. The configuration delegates research and visual creation to researcher, SVG-maker, and Mermaid-maker subagents; it can also run without subagents, with those delegation-based capabilities omitted. It is intended for one learner and is shared as-is under the repository's stated installation instructions.
LiveKit is an open-source framework and developer platform for building, testing, deploying, scaling, and observing real-time voice, video, and physical AI agents. Its agent pipeline streams user speech from an app, browser, or phone call to an agent, which applies custom business logic and returns a response; the platform supports automatic turn detection and interruption handling, speech-to-text, language-model, and text-to-speech providers, web and mobile applications, and telephony through phone numbers and SIP integrations. LiveKit Cloud provides deployment and scaling on LiveKit's real-time infrastructure, alongside an inference gateway and full-stack observability for agent sessions.
LiveKit Agents is an open-source Python framework for building programmable, real-time multimodal voice agents that run as server-side participants. It combines speech-to-text, large language models, text-to-speech, and realtime APIs with LiveKit's WebRTC clients, telephony stack, RPC and data APIs, and MCP tool support. The framework provides agent sessions, server-side job scheduling and dispatch, semantic turn detection based on a transformer model, multi-agent handoffs, and tools for voice, text-only, transcription, vision, and video-avatar applications. Agents can be tested with native test integration, including event assertions and LLM-based judges, and can run locally in console mode, in development mode with hot reloading, or in production mode. The framework can use LiveKit Cloud or a self-hosted LiveKit server and is distributed as a Python package with plugins for model providers. The Agents framework is licensed under Apache-2.0; LiveKit turn-detection models use the LiveKit Model License.
An AI framework for defining model architectures and supporting model serving.
Model Context Protocol (MCP) is an open, standardized protocol layer for connecting large language models and AI agents with external data sources, hosted infrastructure tools, and other contextual data. It defines formats, metadata, protocol schemas, and APIs for sharing information and tools, attaching, referencing, and validating context such as documents, embeddings, and provenance, while supporting scoped authentication and permissions. MCP translates JSON requests from an agent into calls to service APIs, including CRM, container-management, and GKE capabilities, and can provide design-system context and related tools for generating consistent applications. The official project publishes its specification, documentation, and protocol schema; the schema is defined first in TypeScript and also provided as JSON Schema for broader compatibility. The protocol was created by David Soria Parra and Justin Spahr-Summers, is hosted at modelcontextprotocol.io, and is licensed under the MIT License.
The Modular Platform is an open-source platform for AI development and deployment, comprising the MAX Framework and the Mojo programming language. Its repository includes the Mojo compiler and standard library, MAX accelerator libraries and kernels, Python-based model pipelines, code examples, and an OpenAI-compatible MAX inference server. The platform supports model serving and AI development while retaining Python-based workflows. The repository is licensed under the Apache License 2.0 with LLVM Exceptions, while MAX usage and distribution are licensed under the Modular Community License.
Munder Difflin is a free, open-source desktop multi-agent harness that runs multiple terminal-based AI coding agents as coordinated local agents. It wraps agents including Claude Code, OpenAI Codex, Gemini CLI, Qwen, OpenCode, GitHub Copilot CLI, and custom commands, working with users' existing subscriptions and hourly usage limits. Each agent runs as a real node-pty process and is rendered through xterm.js, while an Electron and React interface displays sessions as avatars on a Pixi.js office floor. A GOD orchestrator called Michael assigns and routes work, adjudicates agent messages, and escalates spending, destructive operations, scope changes, and other configured approvals. Agents coordinate through a local Git-backed hive of plain-file memory, atomic mailboxes, a shared blackboard, and an append-only event log; a markdown-first semantic memory layer supports recall across sessions. Optional Git worktrees isolate parallel agents.
Needle 2 is an open-source, 45-million-parameter foundation model for tool calling, device use, and structured extraction on small devices. Its weights are compressed into a single 14 MB binary and integrated with its inference engine; the repository reports that a full session uses about 28 MB of RAM and performs inference without network access after setup. The model converts text and tool descriptions into schema-constrained JSON calls using a byte-level grammar, attaches a learned confidence score for threshold-based escalation, and can retrieve the top five tools from a larger catalogue. It uses a 256-token sliding window with tools pinned as key-value sinks to bound memory, and its Simple Attention Network combines a Hadamard MLP, grouped-query attention, engram key-value memory, and multi-lane hyper-connections. The Python package supports inference, LoRA fine-tuning, and checkpoint export; the inference engine is fetched from Hugging Face and cached, while the training stack is an optional installation.
OCR It is a Chrome and Firefox browser extension for extracting text from paginated document viewers, including scanned books, slide decks, PDFs, and embedded readers where text selection is unavailable. Users select a screen region, capture it with a hotkey, and append the locally generated OCR result to an ordered transcript; an automatic mode captures each page, advances the viewer, and stops when text repeats, paging fails, OCR fails, or a page limit is reached. OCR runs offline through a bundled Tesseract build without outbound requests, API calls, or image uploads. The extension crops, optionally enlarges, and converts the visible region to grayscale, queues OCR jobs serially, stores page text and verification thumbnails, flags consecutive duplicate pages, and exports the transcript as copied text or a TXT file with page separators. Page turning can use a picked screen point or keyboard event, including controls inside supported frames and shadow roots. It uses narrowly scoped browser permissions, relying on activeTab for single captures and requesting an additional site grant for automatic runs that persist across page loads or turn pages inside cross-origin iframes.
An OpenAI API for real-time interactions, powering voice interaction with a 3D globe.
Paseo is a self-hosted platform for orchestrating multiple coding agents, including Claude Code, Codex, GitHub Copilot, OpenCode, and Pi. Its local daemon manages agent processes on users' machines, while desktop, mobile, web, and CLI clients connect to run agents in parallel, stream output, send follow-up tasks, and work in specified directories or worktrees. Paseo supports voice task dictation and control, agent handoffs, advisor and committee workflows, local tools, configurations, skills, and development environments, and provides an MCP server, WebSocket API, and TypeScript SDK for integrations, dashboards, and orchestration services. Remote connections use an end-to-end encrypted relay, TCP, Tailscale, or another VPN. Paseo can run as an installed application, headlessly, or as a Dockerized daemon with a self-hosted web UI. The repository is licensed under Apache-2.0 and states that Paseo has no telemetry, tracking, or forced log-ins.
Proliferate is an open-source AI IDE for running coding agents such as Claude Code, Codex, OpenCode, Cursor, and Grok through their native harnesses. It runs agents in parallel within one workspace, giving each task an isolated Git branch and worktree along with its own terminal, conversation, and review state. Agents can delegate scoped work to subagents, while MCPs, skills, Computer Use, Browser Use, and custom tools can be configured once and shared across agents. Its workflow system supports recurring and event-driven runs such as review passes, alert triage, and dependency updates. The desktop application can use a local runtime or connect to a control plane that runs locally or in the cloud. The control plane is self-hostable through Docker, AWS, GCP, Azure, Kubernetes, or air-gapped deployments. The repository lists Rust, Node.js, and pnpm as source-build requirements and is licensed under AGPL-3.0.
Python is a programming language and the implementation environment maintained in the CPython repository. The videos describe its use for back-end applications, bots, scientific analysis, API scripts, file-operation primitives, database and vector-search result processing, coding-session dashboards, document-converter bindings, and compiled agent skills. CPython can be built on Unix-like systems and Windows, with optional profile-guided and link-time optimization; installable distributions and documentation are provided through python.org.
Rome is an agentic operating environment from Rome OS for collaboration between humans and AI agents. It provides manifests, typed actions, agents, skills, hooks, tools, workflows, memory, interfaces, policies, and persistent database-backed state in a guardrailed environment where agents can build applications, define standard operating procedures, and orchestrate workflows under human guidance. Rome supports one-off tasks, scheduled work, and long-running follow-through. Rome Apps combine a purpose-built web interface, agent reasoning, agent-owned collaborators, reusable workflows and skills, lifecycle hooks, typed actions, HTTP APIs, and persistent app-private databases or files into installable products. Apps are defined with an app.yaml manifest and use the @rome-os/app-runtime and @rome-os/app-web-sdk packages; Rome can generate and iterate on app or workflow source code from a plain-language request. The project can run locally through Docker or be accessed through Rome Cloud, which the repository describes as a private preview environment.
Screendrop is a native, local-first macOS menu bar application for capturing, annotating, recording, editing, compressing, and sharing screenshots and screen recordings. It captures displays, windows, or selected areas; provides configurable hotkeys and after-capture workflows; and includes a floating preview stack, offline History, non-destructive annotation with redaction and mockup effects, and a recording studio with separate screen, camera, microphone, system-audio, cursor, click, and keystroke data. Recording projects support timeline edits, zooms, reconstructed cursors, camera layouts, captions, transcript-based cuts, social aspect ratios, and reusable style presets. On-device speech recognition provides transcription, karaoke-style captions, filler-word removal, silence trimming, and a notch-native teleprompter. Optional self-hosted sharing uses a Cloudflare Worker, R2, and D1: the Worker authenticates uploads, stores media and metadata, and returns share links for screenshot viewers or Loom-style recording pages with transcripts, comments, playback controls, and scrub previews. Captures, projects, annotation sidecars, and transcripts remain on the Mac unless explicitly uploaded, and there is no central Screendrop account or hosted server. The project is distributed as a DMG or Homebrew cask, requires macOS 26.4 or newer according to its README, and is dedicated to the public domain under CC0 1.0 Universal.
SwarmForge is a local, tmux-based orchestration platform for coordinating multiple AI agents across Git worktrees. It reads a project-local configuration that assigns roles, agent backends, worktrees, task or batch handling, and handoff propagation; launches each role in an isolated tmux session; and provides a browser-based pack cockpit for projects, tasks, approvals, clarifications, live status, agent-pane inspection, and teardown. Agents communicate through a daemon-delivered handoff protocol and helper scripts rather than direct tmux messages. The repository supplies two-pack, four-pack, and six-pack workflow templates with role prompts and constitution articles, while projects can define their own swarm topology. It runs locally with zsh, Git, tmux, Babashka, and at least one supported agent backend such as Claude, Codex, Copilot, or Grok; swarm state is kept in the working director
Tailcat is an open-source Go CLI and library from Tailscale that provides Netcat-style connections over Tailscale's encrypted data plane without using the Tailscale control plane or requiring a Tailscale account. A server generates a connection token containing WireGuard and relay metadata; a client uses the token to establish an end-to-end encrypted tunnel, initially through a DERP relay, while magicsock attempts NAT traversal and a direct peer-to-peer UDP connection. Its userspace networking stack does not alter routing tables, DNS settings, or require root access. The CLI can pipe standard input and output between machines, forward local TCP ports, provide an unauthenticated SSH server, run commands through a SOCKS5 proxy, act as an exit node, ping and inspect connection tokens, and publish tokens through DNS TXT records. The Go library exposes server and client APIs for dialing TCP ports through the tunnel. Tailcat can use Tailscale's public rate-limited DERP relays or user-operated relays, and an experimental WebAssembly browser demo supports file and text transfer. The project states that its CLI, API, and wire format have no stability guarantees.
Tailscale is a networking service for creating private WireGuard-based networks between a user's devices and servers. It can provide remote access over SSH and route traffic through an authorized device configured as an exit node, such as a local Mac, so external websites see that device's network address rather than the server's data-center address. The project includes the open-source `tailscaled` daemon and `tailscale` command-line tool, with the daemon running on Linux, Windows, macOS, and to varying degrees on FreeBSD and OpenBSD. The repository also supplies code used by Tailscale's mobile applications; platform-specific GUI wrappers and hosted-service components are separate. Packages are provided for several distributions, and the project uses WireGuard with authentication features including 2FA, OAuth, and SSO.
Tesseract OCR is software for performing offline optical character recognition on captured document regions. Its official documentation includes a user manual and source-code documentation.
vphone-cli is an open-source macOS command-line tool for creating and running virtual iPhones with Apple's Virtualization.framework and PCC research VM infrastructure. It targets Apple Silicon Macs running macOS 15 or later and automates the VM workflow from IPSW download and patching through DFU restore, custom firmware installation, and first boot. The tool supports multiple firmware patch variants, VM creation and management, SSH and VNC connections, IPA installation, and export/import or APFS cloning of VM bundles. Its host control socket provides programmatic screenshots, touch and swipe input, hardware-key events, and clipboard operations, with each action returning an inline screenshot; the repository also describes an MCP server named vphone-mcp for AI-driven end-to-end testing. Operation requires Xcode and the iOS SDK, several Homebrew dependencies, and macOS SIP/AMFI relaxation to permit the required private virtualization entitlements.
A trusted repository where developers discover, install, and publish Claude Code plugins with standardized metadata, MCP configurations, commands, agents, and skills.
An Electron desktop application that coordinates terminal coding agents as a team, giving each agent a pseudo-terminal, memory, mailbox, and role in a shared visual office.
A 14 MB tool-calling model designed for phones, wearables, and other constrained devices, with offline operation, fine-tuning, and confidence-based escalation.
Searchable transcript of Trending Open-Source GitHub Projects : Kaneo, Rome, Paseo, Proliferate & Screendrop #288 — ManuAGI - AutoGPT Tutorials (13:56). Search for a phrase, then click its timestamp to jump straight to that moment in the video.
Captions sourced from the original video on YouTube, published by ManuAGI - AutoGPT Tutorials. The video, its captions and all related intellectual property remain the property of their respective owners; AINotes claims no ownership. Provided for research, accessibility and search — see the Transcript Notice and Copyright Policy.
00:00 Every week, developers release powerful open-source GitHub projects that make coding, automation, and daily workflows faster. This is your weekly GitHub project update video, covering top trending open-source GitHub projects this week. From multi-agent coding tools to self-hosted platforms and local first apps, you'll discover useful and trending developer tools worth adding to your stack.
00:22 Without wasting time, let's get started. Before we jump into today's project updates, here's a quick announcement for everyone. We've launched a brand new YouTube channel called AI Agent Studio, dedicated entirely to AI Agent projects, tutorials, and tools. So, if you're interested in staying up to date with the latest AI agent open source projects, learning how to build your own agents, or exploring cuttingedge agent frameworks, make sure to check it out.
00:51 Subscribe now to get weekly videos, in-depth guides, and realtime project breakdowns. The link is right there in the description. Don't miss it. All right, let's get into today's video. Project number one, Claude Code Plugins directory. A curated marketplace of trusted Claude Code plugins. A curated repository of highquality Claude Code plugins split into internal anthropic builds and external community submissions.
01:16 It solves plug-in discovery by giving each entry a standard structure with metadata. Optional MCP configs/comands, agents, and skills. Developers install any plug-in with one command or browse in app and publish names stay fixed via a rename map. Suits developers extending Claude code. Browse and install a plugin to try it. Project number two, God's Eye View.
01:40 A live 3D globe built entirely from public data feeds. an open-source browser app rendering a photorealistic 3D globe that tracks real aircraft, ships, satellites, earthquakes, traffic, and public cameras. It solves scattered open data by fusing feeds like open sky, AIS stream, Silist, and USGS into one interactive Earth using CesiumJS and Google 3D tiles.
02:04 A voice agent built on OpenAI's real-time API answers questions, annotates, and flies the camera on command. Suits open- source intelligence enthusiasts. Clone the repo and add your keys to start tracking. Project number three, Tailcat. Netcat style tunneling over tail scales network. No account needed. A Go CLI and library working like Netcat but tunneling over Tailscales encrypted plane without its account-based control layer.
02:32 It solves quick encrypted point-to-point connections without signup or routing changes. One side runs a server for a connection token. The other connects with it, upgrading from a derp relay to direct wire guard peer-to-peer. It pipes stdo, forwards TCP ports, runs an off-free SSH server, or acts as a sock 5 proxy. Run the server command and share the token.
02:56 Project number four, agents, a Python framework for building realtime voice AI agents. Livekit agents is an open-source Python framework for real-time multimodal voice agents that see, hear, and understand. It solves the integration burden of speech, language, and telefan by unifying speechtoext, LLMs, texttospech, and real-time APIs in one session with automatic job dispatch, telefan support, and semantic turn detection.
03:24 It works across LiveKit's WebRTC SDKs, supports MCP tools, and includes an LLM judge testing framework, suits builders of conversational voice apps. Install and run the starter agent to hear it talk. Project number five, Modular Platform, a unified compiler and runtime for AI workloads. An open-source repository hosting the Max Framework and Mojo language for AI development and deployment.
03:49 It solves the gap between research code and production infrastructure with one platform spanning a systems level language, a hardware accelerated kernel library, and a model serving runtime. Developers write custom kernels in Mojo, define architectures in Max and serve them via a standard API. Community contributions are welcome. Suits developers and researchers wanting performance without leaving Python.
04:12 Follow the quick start to serve your first model. Project number six, Minder Difflin, a desktop app that turns coding CLS into a coordinated office. An open-source Electron app wrapping terminal coding agents like Claude Code, Codeex, Grock, and Gemini CLI into a coordinated multi-agent team. It solves the chaos of running parallel AI sessions by giving each a real pseudo terminal, memory, and mailbox visualized as characters on a shared 2D office floor with an orchestrator routing and escalating tasks.
04:44 Agents communicate via a gitbacked file system, avoiding conflicts. Suits developers juggling parallel coding agents. Clone the repo and spawn your first agent. Project number seven, Needle 2. A tiny 14 MB model for tool calling on small devices. An open-source tool calling foundation model packaged as a 14 MB binary running full sessions in about 28 MB of RAM.
05:06 It solves structured tool use on tiny hardware like phones and wearables where large models don't fit using a bite level grammar to constrain every token. It runs offline, supports Laura fine-tuning that merges back in and returns calibrated confidence scores for escalation. Suits ondevice AI developers. Install the package and decorate a function to see it call a tool.
05:30 Project number eight, a minimal self-hosted alternative to Jira and Linear. An open-source self-hosted project management platform built with React, Hono, and Typescript. It solves tool bloat by stripping features to essentials, keeping every screen focused on the work. Teams deploy via Docker Compose, a one-click CLI called DRIM, Coolify, or Kubernetes with a Helm chart, keeping data under their own control.
05:57 An official MCP server lets AI tools like Claude manage tasks directly. Suits teams wanting an uncluttered owned canbon tracker. Spin up the compose file and open the dashboard. Project number nine, ego light. One browser shared by you and your AI agents. A Mac OS browser letting a person and their AI agents work in the same app without conflict. It solves the friction of separate automation tools by giving each agent an isolated space with real login and cookies controlled through EGO browser a layer exposing snapshot
06:29 fill click and navigate as JavaScript functions. Multiple spaces run different agents while the person's own tabs stay untouched. Working with clawed code, codecs, and cursor. Install the app and run a task to see it work. Project number 10, Swarm Forge, a T-Mux orchestration layer for discipline multi-agent coding. A lightweight T-Maxbased tool coordinating several AI coding agents working in separate Git work trees on one project.
06:57 It solves agents stepping on each other by giving each role its own work tree, session, and prompt file with a Damon delivered handoff protocol between inboxes. Predefined branches like two pack and six-pack define role chains across claude, codeex, copilot, or grock backends, all bound by a shared rules constitution. Suits developers avoiding config sprawl.
07:20 Pull a runnable branch and run the swarm script. Project number 11, Pacio. One interface to run coding agents from your phone or desk. A self-hosted orchestration platform running coding agents like Claude Code, Codeex, and C-Pilot through one interface across desktop, mobile, web, and CLI. It solves fragmented tool management with a local demon managing agent processes, so dev environments stay intact while clients connect from anywhere.
07:46 It supports voice dictation, parallel work trees, agent handoff skills, and a committee mode for second opinions with no telemetry. Suits developers managing agents remotely. Install the CLI or app and run your first task. Project number 12. Hey Mi multilingual live subtitles that run on CPU alone. An open-source CPUon speechtoext tool generating realtime subtitles with speaker labels and live translation.
08:14 It solves accuracy gaps in general transcription by routing each utterance to a specialist quantized onx model per language. falling back to Meta's omlingual ASR for roughly 1,600 others. No GPU required. Partial subtitles update live finalizing within about a 100 mitzv with speaker tagging and an OBS overlay. Suits streamers needing cloud-free captions.
08:37 Run the quick start script and open the dashboard to transcribe. Project number 13. Vphone CI. Boot a virtual iPhone on your Mac using Apple's own VM tools. an open-source CI that boots a virtual iPhone on Mac OS using Apple's virtualization framework and its published research VM infrastructure. It solves iOS testing without physical hardware via make targets that download IPSW firmware.
09:03 Patch the boot chain, restore it to a virtual disc, and boot the device with an automation socket for screenshots and touch injection. Access runs over SSH, VNC or RPC via USB forwarding. Suits iOS developers needing disposable test devices. Follow the quick start targets to boot one. Project number 14. Screen Drop. A local first menu bar screenshot and recording tool for Mac.
09:30 A native Mac OS menu bar app for capturing, annotating, and sharing screenshots and screen recordings without switching tools. It solves captureedit share friction by keeping everything in one hotkey triggered utility with shapes, blur, and text markup, plus ffmpeg based recording compression. Cloud sharing runs through a self-hosted Cloudflare worker with R2 and D1, keeping files in the user's own account, while local capture history stays offline by default.
09:57 Suits privacy conscious Mac users. Download the DMG or install via Homebrew. Project number 15, Lattis DB, a single file graph database that also does vector and text search. An embedded single file property graph database combining graph traversal, vector similarity, and full text search in one query engine. It solves the need for separate databases by letting one cipher based query touch relationships, embeddings, and text together running with zero configuration via a write- ahead log and single writer model.
10:31 Bindings exist for Python, TypeScript, Go and C, plus durable streams sharing the transaction log suits builders of agent memory or rag pipelines. Open a database file to try it. Project number 16, learn, a personal AI teaching configuration built on PI, a personal AI learning system built as a configuration directory for the PI agent framework, encoding a teaching philosophy as a reusable skill.
10:56 It solves inconsistent AI tutoring by packaging teaching, quiz generation, and diagram creation into skills and sub aents, including a fact-checking researcher and SVG/Merade diagram makers, plus interactive quiz extensions with instant feedback. It can also run without sub aents at reduced capability. Suits Pi users wanting repeatable learning. Clone it into your Pi directory and start a session.
11:22 Project number 17, Rome. An operating system where agents build their own apps, an open-source agentic operating system, giving humans and AI agents a shared environment for tools, workflows, memory, and interfaces. It solves capabilities vanishing after one conversation by letting agents scaffold persistent git tracked apps with their own interface, database, and actions from a plain language request.
11:45 It runs via Docker, hosted Rome Cloud, or a PNPM monorreo with apps sharable through an app store. Suits users wanting reusable agent capabilities. Run the quick start and describe a workflow to see it build. Project number 18, Dupe, a multiplayer canvas where people and AI agents design together. Leave an open-source multiplayer design canvas where people and AI agents build web interfaces together in real time.
12:12 It solves prompt and refresh AI design friction with a built-in MCP server that streams HTML directly into shared canvas frames visible live to everyone via shared cursors and an activity feed. Agents narrate progress and respond to human feedback. Self-hosting via one docker compose command with embedded Postgress suits teams wanting visible agent collaboration.
12:36 Connect an MCP client to watch an agent design. Project number 19. Bans it. Turn a locked viewer document into text page by page offline. A Chrome extension that extracts text from pageionated documents trapped in viewers with no copy function like scanned books or locked PDFs. It solves this by capturing a userdrawn region on hotkey press per page.
12:58 Running fully offline OCR via a bundled Tesseract build and appending results to a transcript with an automatic mode that advances pages and detects duplicates. No data leaves the machine. Suits anyone extracting text from locked viewers. Loaded unpacked and pin a region to start. Project number 20. Proliferate. An open-source IDE for running coding agents in parallel.
13:22 an open-source desktop and web IDE for running coding agents like cloud code, codeex, and cursor in parallel within one workspace. It solves juggling multiple agent tools by giving each its own isolated work tree or cloud sandbox while preserving native O and permissions, sharing MCPs and skills across sessions, showing inline git diffs, and running reviewer agents before humans do. Suits developers coordinating multiple agents. Download the desktop app or run it from source to start a session. Thanks for watching.