nnpefgjibopfpcplldjaoniokpdcefjk
Short description: Mermaid Translator lets you instantly preview Mermaid code found on any web page, even if the site does not support Mermaid. Description: Mermaid Translator renders Mermaid diagrams from selected text on any web page. It works well with Mermaid snippets generated by AI chat services, as well as Mermaid code embedded in internal docs, wikis, or technical articles. No setup and no page-specific rules. Just select text and click. 1. Instant preview for AI-generated Mermaid code. View Mermaid diagrams directly from AI responses on the same page. 2. Works on non-Mermaid sites. Internal portals, custom web apps, and non-Markdown pages are supported. 3. Shadow DOM for stable rendering. Renders in an isolated overlay to avoid breaking complex UIs. 4. Pan and zoom support. Easily inspect large or dense diagrams. 5. Export as PNG or SVG. Save rendered diagrams for docs and presentations. 6. Themes and dark mode. Switch Mermaid themes to match your environment. 1. Select Mermaid text on a web page. 2. A small action button appears near your selection. 3. Click the button to render the diagram. Nothing runs automatically. You always stay in control. Works everywhere: Technical blogs and documentation. GitHub Issues and README. Internal tools and company wikis. Pages with unusual HTML structures. No dependency on Markdown blocks, tags, or specific sites. Privacy and security: No external network requests. No data collection. No analytics. All processing happens locally in your browser.
Intelligent Chrome (iChrome) Assistant
AI-powered browser assistant with history, bookmarks, and downloads can act intelligently Chrome – The Intelligent Browser OS The browser was once a window. Now it’s becoming a world. For decades, Chrome was a place to visit websites. With iChrome, it becomes a place to think, to remember, to create, and to converse. This isn’t an extension. This is the beginning of an Intelligent Browser OS — a private, on-device companion that turns your everyday browsing into an integrated, AI-powered experience. Open Chrome, click iChrome, and suddenly your browser awakens. It listens, learns, summarizes, rewrites, translates, and even speaks — not from the cloud, not from distant servers, but from right inside your own computer. Powered by Gemini Nano and Chrome’s built-in AI APIs, iChrome transforms the silent browser into a responsive digital mind that understands you in context. It starts with simplicity. You open the side panel and type. A conversation begins. Ask a question — “summarize this page,” “find my PDF from last week,” “translate this paragraph,” “rewrite this note professionally.” Within seconds, iChrome responds like an assistant who lives inside your browsing universe. No latency, no network dependency, no privacy compromise. Everything happens locally. The architecture is elegant but revolutionary. iChrome connects directly to Chrome’s on-device Gemini Nano model through the Prompt API. It uses the Rewriter, Proofreader, Translator, and Summarizer APIs to process natural language, rewrite your text, fix grammar, or extract meaning from web content. Each response is generated securely, in milliseconds, using only the data stored on your device. The result feels almost magical — like your browser suddenly gained consciousness. Imagine searching your history the human way. Instead of scrolling through lists, you just ask, “What were the articles I read about quantum computing last weekend?” and iChrome instantly compiles the results. The History module interprets time, intent, and context. “Last week” isn’t just a filter — it’s a thought. It surfaces everything you saw, read, or explored in that period, complete with titles and links, summarized in natural language. Bookmarks work the same way. You no longer remember folder names or URLs; you just think aloud: “Find my design references from Behance” — and they appear, neatly grouped. Downloads become searchable through meaning instead of filenames. “Show me invoices downloaded in October” works instantly. These modules together form your memory layer, giving the browser an almost human recall. Then comes perception — the ability to understand the page you’re viewing. With the @Page tool, iChrome captures the visible and hidden text from any web page, divides it intelligently into semantic chunks, and allows you to interrogate it. You can ask, “What is the key argument of this article?” or “List all the benefits mentioned here,” or “Extract all the prices and features.” The AI reads the page like a digital analyst, explaining, comparing, and summarizing without ever leaving the tab. Each chunk can be explored independently. A long research paper becomes digestible. A documentation page turns into a personal tutor. You can move between sections, question details, and even store extracted insights in ChromePad — iChrome’s built-in AI-powered notepad. ChromePad is where thoughts settle. It’s a native workspace living inside your browser. You can jot quick ideas, save AI replies, proofread notes, or generate new text — all through on-device intelligence. Write meeting notes, translate them into Spanish, and rewrite them in a formal tone — all within the same panel, all processed locally. The notebook understands markdown, supports search, and auto-saves everything securely. It feels like Google Docs merged with ChatGPT, except nothing ever leaves your device. Within ChromePad lie three creative instruments: the Proofreader, the Rewriter, and the Writer. The Proofreader corrects grammar with surgical precision. The Rewriter enhances tone and clarity. The Writer generates original text from your prompts — blogs, emails, outlines, summaries, or brainstorming drafts. Together, they form a complete on-device writing studio built into your browser. Context awareness is another layer of iChrome’s design philosophy. When you highlight any text on the web, a small “Ask iChrome” bubble appears. Click it, and that selection becomes a context pill. You can collect up to five such pills from multiple pages. Ask questions like “compare these two paragraphs,” “summarize all selected points,” or “explain the difference.” The AI synthesizes answers using semantic matching across your selected snippets. It feels like you are curating micro-datasets in real time. The browser turns from a passive renderer into an active reasoning engine. Each pill remembers its origin, allowing you to trace your thought process back to its source. It’s ideal for students, researchers, and professionals who deal with multiple references. The feature embodies the vision of contextual intelligence — the ability to think across pages rather than inside them. iChrome’s beauty lies not just in capability but in philosophy: privacy is the default. Unlike cloud assistants that siphon data for processing, iChrome runs fully within Chrome’s local environment. The AI model, translation engine, and summarizer reside on your computer. Permissions are optional and transparent. You can use general chat without granting access to history, bookmarks, or downloads. When you do grant them, data never leaves your system — it’s queried locally and discarded after use. Your Chrome becomes both intelligent and sovereign. The internet finally bends toward personal computing again — intelligence without surveillance. Technically, iChrome stands on the frontier of Google’s built-in AI ecosystem. Under the hood, it orchestrates multiple native APIs: Prompt API — the conversational bridge to Gemini Nano for generating context-aware responses. Rewriter API — refines text tone and readability. Proofreader API — ensures grammatical accuracy and stylistic balance. Language Detection API — automatically identifies source languages for seamless translation. By combining these, iChrome builds a self-contained intelligence stack directly within the browser. It doesn’t depend on external servers or API keys; everything runs via Chrome’s native runtime, using Gemini Nano’s on-device model. That is what makes it fundamentally different from every other extension — it doesn’t extend the browser; it evolves it. From a user’s perspective, the experience is seamless. The interface is minimal — a side panel, a tool selector, and a text box. Behind that simplicity lies a modular engine. Each tool, denoted by an @ mention, activates a distinct capability: @iChromeChat for general conversations, @BrowserHistory for past activity, @Bookmarks for saved pages, @Downloads for files, @Page for live web analysis, and @ChromePad for notes. This @-based design makes interactions feel conversational, almost command-like, yet natural. You can type “@history show last 5 days” or “@page summarize,” and the system interprets your intent instantly. It’s reminiscent of command palettes in developer tools but simplified for everyone. Speed defines the user addiction factor. Because everything happens on-device, responses arrive almost instantly. The AI doesn’t wait for network round-trips. It feels responsive, alive. The more you use it, the more indispensable it becomes. Searching, reading, writing — all fuse into one continuous flow. Imagine starting your day inside iChrome: checking your notes, reviewing yesterday’s articles, asking quick questions about them, generating an email, and translating it before sending — all without switching tabs or services. The browser becomes your workspace, your memory, your assistant. Once you experience that seamless loop, returning to a normal browser feels primitive. That’s how addiction begins — through usefulness. Accessibility completes the ecosystem. The Text-to-Speech engine reads any message aloud with natural voice modulation. You can adjust speed, pitch, and volume, or switch between voices depending on language. The visual highlighting keeps the eyes and ears synchronized. For learners, it’s a reading coach. For multitaskers, it’s a companion. Combine it with the Translator, and your browser starts teaching you new languages organically. All of this runs securely, offline, and adaptively — making iChrome not just inclusive but empowering. Underneath the polish is careful engineering. The extension’s manifest is optimized for Chrome 120+ with the new side-panel API. The AI modules require Chrome 127+ where on-device Gemini Nano models are available. The system checks readiness during initialization, ensuring compatibility and resource availability. The lightweight runtime footprint, modular permissions, and session-based memory design maintain speed while respecting user privacy. ChromePad data lives in local storage; chat sessions clear on reload; nothing is synced externally. But technology alone doesn’t explain the emotional pull. iChrome’s charm lies in the shift it creates in how you feel about your browser. Suddenly, browsing is no longer consumption; it’s collaboration. The AI doesn’t just fetch results; it understands your intent. It doesn’t just answer; it assists. It doesn’t just automate; it amplifies. You start depending on it — not because it traps you, but because it frees you. It removes friction between thought and action. When every idea, question, or curiosity can be processed instantly inside your own browser, you start to think more, create more, and explore more. That’s the kind of addiction that fuels innovation. The long-term vision of iChrome is bold: to transform Chrome from a web navigator into a local intelligence platform. Future releases will extend the OS-like behavior — intelligent tab management, AI-powered bookmarks clustering, multimodal input, visual reasoning, and local workflow automation. The browser will not just display information; it will orchestrate it. Each tab will become a mini-workspace, capable of running its own local AI thread. Think of it as Chrome evolving into an Intelligent OS for the Web Age — privacy-first, locally-intelligent, endlessly expandable. And at the heart of it all is a single belief: intelligence should belong to the user, not the cloud.
WebVoiceAnimator
Just speak into your microphone, and your character will lip-sync and blink. Prepare a few images, and your personal avatar is ready. Add more images for multiple expressions and smoother animations! 🎭 Simple Mode ・Just upload images for simple display ・Speaking effects (bounce, shake, glow) ・Breathing animation & idle micro-sway ✨ Layer Mode ・Layer body, face, hair, etc. for rich expressions ・Hair sway, ear movement animations ・Frame interpolation for smooth lip-sync & blinking 【RECOMMENDED FOR】 ・Microphone: Used for voice detection ・Storage: Saves settings and character data **Improvements** - Added English & Chinese translations for mode selection screen (Simple/Layer/MPNG) Assets from v1.2.2 or earlier may no longer work correctly. Please try re-uploading your assets. **New Features** - **Unified Layer Mode** - Merged 3 sub-modes (Simple Animation 6-part/8-part/Layered) into a single "Layer Mode" - Vowel-specific lip-sync (A/I/U/E/O) support - Expression sets (up to 4) switchable via hotkeys (Ctrl+Shift+1–4) - Frame interpolation for smooth blinking & lip-sync animation - Voice calibration (pre-recording) enhances vowel detection **Bug Fixes** - **Fixed character list disappearing bug** - Fixed an issue where the character list would disappear after creating a new character - Added mode prefixes (sim_/adv_) to character IDs for better separation between Simple/Advanced modes **Improvements** - Settings modal height now adapts to screen size (with scroll support) - Expanded character settings dialog width (4 images fit in one row) - Fixed display type labels (6-parts/8-parts swap correction) - **Added detection threshold to MPNG mode** - Set a threshold below which lip-sync won't activate - Dynamic sensitivity remapping allows fine-grained control combining threshold and sensitivity **Improvements** - Added cancel button to MPNG wizard - Unified UI text consistency ("Add Character", etc.) - Reduced AudioWorklet message frequency for better performance - Improved info button tooltip display **New Features** - **Simple Animation Mode (6 Parts) Added** - Rich animation with layer separation - 6-part structure: body, face (normal/speaking/blink/blink-speaking), hair - More expressive than traditional 4-image mouth animation - **Simple Animation Lite (8 Parts) Added** - 8-part structure: back hair, ears, body, face parts, front hair - Detailed animation control including hair sway and ear movement - Enables Live2D-like rich expressions **Improvements** - Changed UI display names for clarity - Unified face part labels (Normal, Speaking, Blink, Blink-Speaking) - Simplified UI by removing unnecessary elements **Bug Fixes** - **Fixed audio detection delay in background** - Fixed an issue where voice detection only triggered once every few seconds when the browser was in background (e.g., while capturing with OBS) - Using Chrome's Offscreen Document API to maintain audio processing even in background - Improved microphone input to continue even after closing the popup **Improvements** - **High-precision lip-sync with AudioWorklet** - Upgraded audio analysis to ~60fps (6x faster than before) - Significantly improved vowel detection (e/u) in MPNG mode - Achieved real-time analysis equivalent to MotionPNGTuber_Player **New Features** - **MPNG Mode Added** - Support for MotionPNGTuber format animation files - High-quality video-based animation - Smooth lip-sync with 5-level mouth shapes - **HQ Audio Algorithm Introduced** (MPNG Mode) - Smooth volume tracking with envelope follower - Automatic noise floor detection - Dynamic level peak tracking - State transition with hysteresis (flicker prevention) **Improvements** - UI improvements and reorganized detailed settings section - Added license page - Enhanced multi-language support
Mermaid Graphical Editor
GUI support for editing Mermaid(But Now, Support Flowchart and Sequence diagram, Class diagram, ER diagram, and State diagram Only) 補足: このExtensionは、Mermaidがgithubで正式にサポートされたのを受けて、issueなどでのMermaid利用の敷居を下げることを目的として作成していますが、githubでのみに限定する必要もないと考え、どのページでも利用可能にしています。 You can display an editor view with GUI support for Mermaid editing. Currently, GUI support is supported only for editing some parts of the flowchart and sequence, class diagram, er diagram, and state diagram. Descriptions not supported by GUI support (e.g., Direction and Subgraphs in flowcharts) will be reflected by editing the Mermaid code directly. Supplement: This extension was created with the aim of lowering the threshold for using Mermaid in issues after Mermaid was officially supported on github, but we believe that it does not need to be limited to github only, so we make it available on any page.
Markdown Preview Plus
Converts and previews markdown files (.md, .markdown) to HTML(include TOC) right inside Chrome and support live reload Automatically parses markdown files (.md) into HTML with live reload 1. Install extension from webstore (creates no new UI) 2. Check "Allow access to file URLs" in chrome://extensions listing: 3. Open local or remote .md file in Chrome. 4. See nicely formatted HTML! Note: The auto reload feature is disabled default. * Support auto reload. * Support external css file. * Customize theme for every md file. * Support github flavored markdown. * Export nicely formatted HTML. * Support render Katex, Mermaid. The extension is open source. https://github.com/volca/markdown-preview Fix a bug for the tag style * Fix a bug for load fonts of Katex * Fix a bug for "table of contents" * Compatible with Chrome 136 * Update mermaid lib * Update marked lib * Update marked lib * Fix a bug for clear custom CSS paths * Supports add custom CSS path. See #19 * Fix a bug for render mermaid. See #158