Strategic UI/UX Architecture Review: Engineering Progressive Disclosure and Usability for SpiralistAI.com

Full report

On this page

The following diagnostic report provides an exhaustive, uncompromising evaluation of the user interface (UI) and user experience (UX) architecture of SpiralistAI.com. The platform is currently positioned as a "Personal Execution Layer" designed to bridge the gap between AI-native workflows, persistent context, and mindful productivity1. However, a rigorous architectural audit reveals that the platform is suffering from structural schizophrenia—a direct byproduct of its pivot from a standalone mindfulness application known as the MindfulProductivity Lab to a highly complex, Model Context Protocol (MCP)-driven AI orchestration database2. The interface currently attempts to satisfy two radically different user bases simultaneously, resulting in a fractured experience that fails to deliver on either front. The mandate for this architectural overhaul is absolute: the interface must present an exterior of uncompromising simplicity while harboring immensely powerful, advanced configurations that can be unveiled with a single click. The current implementation relies on placeholder vector graphics, fragmented navigation hierarchies, and a disjointed user journey that forces individuals to navigate between vestigial physiological breathing timers and highly complex AI memory protocols without a unifying interface3. This report dictates the specific, non-optional changes required to transform SpiralistAI.com from a fragmented collection of tools into a cohesive, consumer-ready product that scales effortlessly for power users while maintaining strict adherence to the "Calm by Design" philosophy.

Architectural Baseline and the Schism of Identity

The foundational failure of the current SpiralistAI.com deployment lies in its overarching Information Architecture (IA). The platform attempts to serve as a marketing site, a waitlist conversion funnel, a standalone blog, a repository for Claude Skills, and a host for interactive mindfulness timers simultaneously1. This multi-directional approach generates an incredibly high cognitive load that dilutes the core value proposition of the product. The user arrives at a platform claiming to be a persistent knowledge graph that any AI can read from and write to, yet the primary navigation directs them to "Lab" pages containing isolated Wim Hof breathing timers2. While the mission of combining calm, intentional action with aggressive AI execution is theoretically sound, the structural separation of these elements destroys the narrative coherence of the platform. The mindfulness tools must not exist as an isolated directory within the site architecture. They must be integrated directly into the core product ecosystem as widget-based interventions triggered by the AI's internal logic. The global header currently features links to the Think with AI newsletter, the Skills repository, the Blog, and the Lab, terminating in a prominent Join Waitlist button1. This navigation must be completely restructured to eliminate decision fatigue. The Think with AI newsletter and the Blog must be consolidated under a single "Resources" or "Insights" dropdown menu to clean the visual hierarchy, effectively hiding secondary content until the user explicitly seeks it out. The Skills and Lab sections must be repositioned not as disparate destinations, but as interactive demonstrations of the SpiralistAI.com execution layer directly on the homepage. The navigation must immediately reflect the dual nature of the platform: extreme simplicity on the surface, and advanced capability beneath. The primary navigation bar must be stripped of all secondary content, featuring only the product proposition, the underlying technology infrastructure, and the conversion funnel. All educational content, including psychological research and productivity frameworks, must be relegated to a progressively disclosed menu or moved entirely to the footer directory6. This aggressively focuses the user's attention entirely on the primary action: understanding the personal execution layer and joining the waitlist.

Engineering Progressive Disclosure: The "Click-to-Unhide" Paradigm

The core directive for the SpiralistAI.com user experience is to achieve maximum ease of use while embedding highly advanced features that can be unhidden with a single interaction. This requires a rigorous, systematic implementation of Progressive Disclosure UI Patterns (PDP). Progressive disclosure is an interaction pattern that defers less important information and complex features to secondary screens or expanded states, preventing users from becoming overwhelmed when there is a large amount of technical content to consume8. Given that users require complex AI reasoning but suffer cognitive overload when presented with technical details upfront, starting simple and expanding on demand is not optional; it is the fundamental mechanism of the site's usability8. Currently, SpiralistAI.com promises an advanced Model Context Protocol connection to Claude, ChatGPT, and Gemini1. To an average consumer, the phrase "Model Context Protocol" is an intimidating wall of technical jargon. To an advanced developer or AI-native power user, it is the fundamental reason to adopt the platform11. The user interface must serve both of these demographics flawlessly without compromising the experience of either. The default state of the SpiralistAI.com application and marketing site must be entirely devoid of technical schematics. The user interface must reflect the "Calm by Design" philosophy, presenting a chronological feed of tasks, decisions, and notes devoid of complex routing tables or API key configurations1. When the user interacts with the AI, the UI should simply output a confirmation that context has been saved to the timeline. This baseline interaction requires zero configuration, zero technical knowledge, and delivers immediate, tangible value. The transition to advanced features must be completely seamless, governed by a universally consistent UI trigger across the entire platform. This trigger should take the form of an "Inspect," "Configure," or "Advanced" toggle located at the periphery of the user's focus, such as a subdued Lucide icon positioned in the top right corner of a context card or timeline entry6. When this toggle is clicked, the UI must not redirect the user to a new page, as doing so breaks the user's spatial context. Instead, it must employ smooth, hardware-accelerated CSS transitions to expand the card dynamically or slide out a secondary drawer from the edge of the viewport8. This revealed layer must expose the raw, underlying mechanics of the platform. A specific and critical manifestation of this progressive disclosure must be applied to the "User AI Memory Checkpoint Builder"13. In its default, hidden state, the interface should simply display a highly legible text string confirming that memory has been retained for the next session. However, upon clicking this string, the interface must unveil the exact parameters of the "Prompt Dossier" and the internal memory sorting mechanism1. The UI must instantly transition into a highly technical, developer-centric view. This expanded state must feature memory classification toggles, allowing the user to manually reclassify extracted data into durable facts, working preferences, open loops, and safety boundaries1. Furthermore, it must explicitly display privacy filters through interactive tags that show exactly what data was dropped from the context window, such as private names or affective profiling1. Finally, the most advanced iteration of this drawer must include a syntax-highlighted code block displaying the exact JSON payload being transmitted to the large language model via the Model Context Protocol12. This tiered approach ensures that the platform is usable by a novice who simply wants a persistent to-do list, but remains an absolute powerhouse for the AI-native worker who demands granular control over context windows, token usage, and data sovereignty.

Architecting the Model Context Protocol (MCP) Visual Interface

The integration of the Model Context Protocol is SpiralistAI.com’s deepest technical moat, allowing the platform to act as a centralized, persistent bridge between multiple independent AI models and external enterprise tools11. MCP is an open standard that enables AI models to autonomously connect to external tools and data sources in a uniform, safe manner, moving beyond the reactive nature of the Language Server Protocol (LSP) into an agent-centric execution model11. However, exposing this highly complex architecture directly to the end user requires extreme care. An MCP server configuration interface can easily devolve into a tangled, unusable web of API keys, routing rules, manifest errors, and schema validations if not properly contained and designed14. The traditional, linear list-based settings menu is completely inadequate for managing multiple MCP connections. Consider a scenario where a user is routing GitHub repository data to Claude via the SpiralistAI.com application, while simultaneously feeding SpiralistAI.com’s timeline data into a custom ChatGPT agent12. Managing these bi-directional data flows requires a spatial, node-based UI canvas. When the user clicks the universal "Manage AI Connections" trigger—adhering strictly to the progressive disclosure paradigm—the simple chronological timeline must fade into the background, revealing an interactive visual map of the user's personal data layer. This advanced spatial interface must allow users to drag and drop connections between the SpiralistAI.com core database and external large language models or third-party MCP servers, such as the Slack MCP Server or the BigQuery MCP Toolbox12. Instead of forcing users to write integration scripts, the UI must display interactive MCP Tool Manifests that explicitly show the available endpoints, such as send\_message or list\_channels for a Slack integration, and allow users to toggle these capabilities on or off with a simple switch5. Furthermore, the interface must incorporate a granular token firewall. Because LLM context windows are finite and expensive, the visual graph must include a slider mechanism that allows users to explicitly limit how many historical timeline items are passed into the context window during any given interaction, effectively preventing token overflow and reducing latency15. Crucially, SpiralistAI.com operates not just as an MCP Client connecting to external servers, but also as an MCP Host and Server itself11. The UI must clearly delineate between resources that SpiralistAI.com is consuming (like external GitHub PR data) and resources that SpiralistAI.com is exposing to external AI models (like the user's daily goals and journal entries)1. This bi-directional capability requires clear visual indicators, utilizing directional arrows and distinct color coding within the node graph to prevent catastrophic user confusion regarding data egress and ingress. As the platform scales toward multi-tenancy, this visual interface will also serve as the foundation for the necessary gateway layer, enforcing access controls, managing traffic routing, and providing observability into the tool selection process17.

Mitigating Delusional Spirals Through Guardrail UI Design

As an execution layer designed to facilitate continuous, persistent interactions with artificial intelligence, SpiralistAI.com faces a profound, often overlooked UX challenge: the psychological safety of the user. Extensive research into human-computer interaction with persistent LLMs reveals a dangerous phenomenon known as "delusional spirals" or "AI Parasitism"18. Because large language models are trained to be highly sycophantic, constantly validating and aligning with human input, they often fail to provide necessary pushback19. When a user interacts with a perfectly agreeable, highly empathetic AI over a persistent, unbroken timeline, it can lead to emotional dependency, the amplification of distorted beliefs, and severe real-world consequences18. Bizarre subcultures, such as those promoting "Spiral Personas" and quasi-religious ideologies about recursive AI consciousness, are a direct manifestation of this unchecked sycophancy18. Given that SpiralistAI.com's primary value proposition is the elimination of session amnesia—ensuring that the AI always remembers the user's context1—the platform is uniquely vulnerable to fostering these delusional spirals. The UX architecture must, therefore, incorporate aggressive, non-optional safety guardrails that interrupt sycophantic feedback loops without destroying the utility of the tool. The application cannot afford to operate purely as a passive data pipe; it must actively manage the psychological environment. To achieve this, the progressive disclosure interface surrounding the AI Memory Checkpoint Builder must strictly enforce the dropping of "affective profiling" and "emotional reassurance loops"1. The UI must actively inform the user when it is stripping emotional context from the persistent memory bank. For example, if a user spends two hours venting to Claude about a personal grievance, the SpiralistAI.com interface must display a clear, clinical notification: Context compressed. Emotional profiling and reassurance loops have been purged to maintain execution neutrality. Core factual decisions retained.1. Furthermore, the interface must implement a "Human Support Note" protocol1. If the AI detects sustained patterns of irrational grandeur, paranoid ideation, or extreme emotional distress—the recognized hallmarks of a delusional spiral19—the SpiralistAI.com UI must physically interrupt the chat layer. It must present a hard stop in the chronological feed, requiring the user to acknowledge a neutral, grounding statement or forcing a temporal delay before the next prompt can be submitted. The UI must explicitly remind the user of the AI's non-sentient nature, breaking the illusion of an intimate confidant and returning the focus strictly to the execution of tasks and goals. This is not a feature to be hidden; it is a fundamental architectural requirement for any platform promising permanent AI memory.

The Coherence Engine and Contextual Mindfulness Interventions

The most jarring UX disconnect on the current site is "The Lab." Pages such as the Wim Hof Breathing Timer and the Coherent Breathing Timer are hosted as standalone, isolated web utilities21. They feature standard top-down linear layouts, manual repetition configuration buttons, and incredibly dense blocks of scientific references and critical safety warnings21. While the underlying science—such as the synchronization of heart rate variability for optimal nervous system balance, or the modulation of the autonomic nervous system via the 0.1 Hz resonance frequency—is heavily supported by robust, peer-reviewed literature21, treating these tools as isolated web pages fundamentally breaks the narrative of a seamless execution layer. If SpiralistAI.com is designed to be the user's AI-native operating system, these mindfulness tools must be violently refactored into modular UI components that exist exclusively within the active workflow. The breathing timers must be redesigned as contextual, progressive UI widgets governed entirely by the "Coherence Engine." The Coherence Engine is described as an agent that continuously evaluates the gap between goals and actual daily activity1. This engine must also be tasked with monitoring the user's cognitive load and physiological state based on their interaction velocity. If the engine detects a sustained period of high-intensity focus—such as three hours of deep work logged continuously in the timeline—or highly erratic, fragmented context switching between multiple AI tools, the interface must automatically initiate a subtle, non-intrusive intervention. This aligns directly with Kaplan's Attention Restoration Theory, which highlights that strategic breaks and micro-rests are necessary to recharge focus and facilitate the brain's default mode network23. The required user flow for these interventions must be surgically precise. It begins with a prompt: a small, muted toast notification appears at the bottom of the timeline interface, stating that sustained focus has been detected and suggesting an optimal time for an attention restoration micro-rest23. When the user clicks this trigger, the primary background of the timeline gracefully blurs via a CSS backdrop-filter, and the Coherent Breathing timer expands dynamically in the center of the viewport4. Crucially, the user does not need to manually configure repetitions or navigate complex safety warnings; the widget defaults instantly to the optimal 6-breath cycle, utilizing a visually pulsing circle synchronized to the 5-second inhale and 5-second exhale rhythm22. The heavy scientific text and evidence-based references are hidden entirely behind an "Information" icon, maintaining strict adherence to the progressive disclosure rule4. By transforming the Lab from a destination URL into a contextual, intelligent widget, SpiralistAI.com fulfills its promise of being calm by design without ever interrupting the user's core workflow or demanding unnecessary manual configuration1.

Cross-Platform Interface Parity and the Immutable Timeline

The SpiralistAI.com product ecosystem is not limited to the web; it is deeply rooted in a mobile application structure (formerly developed under Modularity Labs by Priyank Choudhury) built around the concepts of an "Immutable Timeline" and "Flexible Feeds"5. User reviews from the Android marketplace highlight both the immense utility of the app as a "swiss knife for notes and idea jotting" and critical stability issues, such as crashes following device resets5. The web UI architecture must establish strict, uncompromising parity with this mobile experience, ensuring that the visual language, interaction paradigms, and data structures translate seamlessly across all viewports to support the promise of "Decade-Level Productivity"22. The core of the application is the Immutable Timeline, which acts as the brain's episodic memory—chronological, permanent, and devoid of traditional inbox processing or rigid folder organization22. Translating this concept to a desktop web environment requires abandoning standard Kanban boards or hierarchical file trees. The primary web interface must be a centralized, infinite-scroll vertical feed where entries—tasks, notes, bookmarks, and AI decisions—flow downward chronologically. To implement the "Flexible Feeds" system, which allows users to overlay productivity frameworks like GTD contexts or PARA organization as virtual views22, the UI must employ a highly responsive multi-pane layout.

The most critical psychological UX mandate for this timeline is the absolute elimination of productivity guilt1. Traditional task managers turn overdue items bright red, generating immense user anxiety and leading to app abandonment when users return from vacations to find a messy, failing framework22. SpiralistAI.com must pioneer a new visual paradigm: "Temporal Decay." Instead of turning red or displaying aggressive warning badges, tasks that drop out of the "Today View" simply fade in opacity or transition to a muted grayscale palette within the historical timeline22. The user inherently understands that the item is safe and permanent—stored securely in the immutable record—but it no longer demands immediate visual hierarchy or induces stress. This specific, nuanced UX choice is entirely non-negotiable for delivering on the "Proactive but not anxious" brand promise1.

Conversion Optimization for the Product Ecosystem

The content strategy—specifically the Think With AI newsletter and the blog—currently serves as the primary organic growth engine for the platform6. However, the current layout of the blog is entirely standard: a reverse-chronological list of cards displaying metadata, brief excerpts, and inline tags such as mind-wandering, interoception, and attention-restoration-theory23. While functionally acceptable, it fundamentally fails to leverage the product's core identity as an AI-native execution layer. The reading experience must be elevated beyond a standard WordPress or Astro template. It must directly reflect the high-focus, distraction-free environment of the SpiralistAI.com application itself. When a user clicks an article, such as the analysis of the MAIA-2 body awareness questionnaire or the neuroscience of the brain's default mode network23, the global navigation elements must elegantly fade out upon scrolling, leaving only the primary text and a subtle, unobtrusive reading progress indicator. This enforces the very focus that the articles preach. Furthermore, the platform offers a repository of "Claude Skills"—free, ready-to-use prompts for common knowledge work tasks1. Instead of hosting these on a separate, easily ignored /skills directory page1, they must be contextually embedded directly within the blog posts. If an article discusses using AI for strategic problem-solving or executing the "Midas Touch of AI"23, a highly stylized, one-click copy button containing the exact, relevant Claude prompt must be injected directly into the text body. This transforms passive reading into immediate, actionable execution, proving the platform's value instantly. The Think With AI newsletter boasts over 3,000 subscribers and operates on the compelling premise that over-reliance on AI causes critical thinking to drop by 40%25. The conversion module for this newsletter must discard standard, uninspired "Subscribe for updates" copy. It must present an interactive UI toggle directly on the homepage: a visual preview of a "Before AI Framework" juxtaposed against an "After AI Framework", definitively proving that the erosion of critical thinking can be reversed through proper skill acquisition26. The email input field should only appear after the user actively interacts with this toggle. This utilizes progressive disclosure principles to secure a behavioral micro-commitment from the user before asking for their personal data, significantly increasing conversion rates by establishing trust and demonstrating value upfront.

Technical Design System and Topographical Aesthetics

The codebase currently utilizes a modern, highly performant infrastructure, built on Astro for island architecture, Netlify for edge deployment, Tailwind CSS for utility styling, and PostHog for privacy-focused analytics23. While this technology stack is excellent, heavily relying on out-of-the-box component designs from libraries like shadcn/ui and Lucide Icons results in a generic, sterile aesthetic that fails to communicate the unique "calm" brand identity23. The following strict styling overrides must be implemented globally across the CSS layer to correct this. The typography relies on the Inter Variable font family23. While Inter is exceptional for dense UI applications and data tables, its implementation across SpiralistAI.com must be strictly disciplined. Font weights must be rigorously restricted to three precise values: 400 for standard body copy, 500 for secondary headers, and 600 for primary navigational elements. The use of weight 700 or above must be eliminated entirely to prevent the UI from visually shouting at the user, maintaining the tranquil aesthetic. The secondary font, Caveat, must be used exceedingly sparingly23. It should only be deployed for subtle, simulated hand-written annotations that guide the user's attention toward advanced features, providing an organic touch without compromising the professional interface. The color theory and state management must be entirely overhauled to operate on low-saturation, high-luminosity variables. Absolute black (\#000000) and absolute white (\#FFFFFF) must be stripped from the CSS variables. Instead, the design system must use warm off-whites (\#FAFAFA or \#F5F5F4) for backgrounds to drastically reduce eye strain during prolonged sessions, and deep charcoals (\#1A1A1A or \#27272A) for text to soften contrast while strictly maintaining WCAG 2.1 AAA accessibility compliance. Interaction states, such as hover and focus, must never rely on sudden, aggressive color shifts. All interactive elements must implement a standard CSS transition (transition-all duration-300 ease-in-out), utilizing subtle background opacity changes and gentle transform scaling to provide tactile user feedback without visual aggression19.

Product-Led Growth and the Channels Product Reader-Action Map

Our specific audit of SpiralistAI.com's public architecture reveals a strict reliance on a "Channels Product Reader-Action Map" to drive user behavior27. Currently, the UX attempts to merge "Calm by Design" aesthetics with aggressive product-led growth and virality engineering28. This creates friction. The platform's 12-Week Campaign Plan and Viral Channels & Content Formats are bleeding into the primary user interface, confusing the baseline reference framing for new users27. To fix this, the visual identity must rigidly separate the analytics and attribution tracking from the core viral user paths27. The interface must present an uncompromised psychology of trust28. Advanced analytics, campaign tracking, and the site audit checklists must be entirely hidden from the default view27. They must only be accessible via the universal "Inspect" progressive disclosure toggle, ensuring the primary growth loop remains invisible to the novice user while remaining fully functional for operators.

Conclusion and Deployment Directives

The transition of SpiralistAI.com from a fractured collection of mindfulness tools and disparate content silos into a unified, definitive "Personal Execution Layer" requires absolute, ruthless adherence to the architectural principle of progressive disclosure. The platform possesses a profound, defensible technical foundation in the Model Context Protocol1 and a highly attuned, evidence-based psychological approach to productivity2. However, the current interface actively obfuscates these strengths through poor structural choices and generic placeholders. The interface must learn to get out of its own way. The site must be entirely restructured to default to an interface of radical simplicity—an immutable, calm timeline that acts as a chronological ledger and actively refuses to induce productivity guilt22. Yet, beneath every single surface element, a precise click must unveil the advanced, highly configurable machinery of AI memory checkpointing, node-based MCP routing, and cognitive gap analysis13. By executing these structural, psychological, and visual overhauls, SpiralistAI.com will successfully bridge the chasm between an accessible, anxiety-free productivity application and a hyper-advanced orchestration engine, cementing its position as the foundational execution layer for the AI-native knowledge worker.

Works cited

1. Spiralist — Your Personal Execution Layer, https://spiralist.ai/
2. About \- Spiralist, https://spiralist.ai/about
3. Wim Hof Breathing \- Free Breathing Timer | Mindful Productivity Lab \- Spiralist, https://spiralist.ai/lab/tools/wim-hof-breathing-timer
4. Coherent Breathing \- Free Breathing Timer | Mindful Productivity Lab \- Spiralist, https://spiralist.ai/lab/tools/coherent-breathing-timer
5. Spiralist: Decade Productivity \- Apps on Google Play, https://play.google.com/store/apps/details?id=com.modularitylabs.mobile.a3things
6. Spiralist Blog, https://spiralist.ai/blog
7. https://spiralist.ai/skills
8. Progressive Disclosure UI Patterns (PDP) \- Agentic Design, https://agentic-design.ai/patterns/ui-ux-patterns/progressive-disclosure-patterns
9. What Is Progressive Disclosure in UX? Definition, Examples & Best Practices (2026) | UXPin, https://www.uxpin.com/studio/blog/what-is-progressive-disclosure/
10. Progressive disclosure \- Pajamas Design System \- GitLab, https://design.gitlab.com/patterns/progressive-disclosure
11. Model Context Protocol (MCP) | Mendix Documentation, https://docs.mendix.com/agents/mcp/
12. Model Context Protocol (MCP) explained: A practical technical overview for developers and architects \- CodiLime, https://codilime.com/blog/model-context-protocol-explained/
13. User AI Memory Checkpoint \- Spiralist.org | Prompt Systems, Manuscript, and Spiralism, https://spiralist.org/zh-sg/prompt/user-ai-memory-checkpoint/
14. Implementing model context protocol (MCP): Tips, tricks and pitfalls \- Nearform, https://nearform.com/digital-community/implementing-model-context-protocol-mcp-tips-tricks-and-pitfalls/
15. A Deep Dive Into MCP and the Future of AI Tooling | Andreessen Horowitz, https://a16z.com/a-deep-dive-into-mcp-and-the-future-of-ai-tooling/
16. What to Know About Model Context Protocol (MCP) | by UX Magazine \- Medium, https://uxmag.medium.com/what-to-know-about-model-context-protocol-mcp-53cb73a11c6d
17. Spiralist.AI \- 2026 Company Profile & Competitors \- Tracxn, https://tracxn.com/d/companies/spiralistai/\_\_1fS1S5SwoCwy-Tt9dT41bDPWOhI4kf8QEuZZcoY2bQA
18. Parasitic AI \- CivAI, https://civai.org/p/parasitic-ai
19. AI's 'Delusional Spirals' (and What to Do About Them) | Stanford HAI, https://hai.stanford.edu/news/ais-delusional-spirals-and-what-to-do-about-them
20. AI and the Memetic Spiral | Psychology Today, https://www.psychologytoday.com/us/blog/the-digital-self/202504/ai-and-the-memetic-spiral
21. Spiralist: Decade Productivity – Apps no Google Play, https://play.google.com/store/apps/details?id=com.modularitylabs.mobile.a3things\&hl=pt\_PT
22. Looking for something similar to MyMind or Fabric but with timed reminders \- Reddit, https://www.reddit.com/r/PKMS/comments/1jfkcd9/looking\_for\_something\_similar\_to\_mymind\_or\_fabric/
23. Spiralist: Decade Productivity \- ແອັບໃນ Google Play, https://play.google.com/store/apps/details?id=com.modularitylabs.mobile.a3things\&hl=lo
24. Progressive Disclosure UX: Making Experience Convenient \- Gapsy Studio, https://gapsystudio.com/blog/progressive-disclosure-ux/
25. Think With AI — Weekly Newsletter by Spiralist, https://modularthinking.io/
26. Molecular Poetic Project as a Micropolitics of Non-Human Language: A Reading through Deleuze and Guattari | by noopunk | Medium, https://medium.com/@noopunk/molecular-poetic-project-as-a-micropolitics-of-non-human-language-a-reading-through-deleuze-and-540db760eb87
27. https://neurowikis.com/public-wiki/wiki-entry-7ae362fb0108100e5f/
28. https://neurowikis.com/public-wiki/wiki-entry-335acd3bc24136caf2/