ULTRON Core

Cinevva's analyzer read this game's design doc and its source, then wrote the brief below. "Make your own" hands it to the game creator as your opening message, so you start from a real spec rather than a one-line idea.

Rebuild “ULTRON Core” as an interactive 3D knowledge-interface experience, not a conventional combat game. Use a black full-screen background with a Three.js-style perspective camera aimed at the origin. In the center, render a bright white faceted icosahedral core/knowledge atom with a soft pale-blue translucent halo. Surround it with six thin elliptical orbital rings at different tilts and radii; they should slowly rotate at different speeds. Place 48 small glowing white knowledge balls around those orbits, with faint blue-white connecting lines forming a graph. The atom and its nodes should subtly pulse and rotate continuously. Add a thin circular 3D voice waveform around the core whose color and activity respond to the current interface state, plus expanding translucent activity rings while listening or speaking.

The camera must support mouse/touch drag to orbit around the central atom and wheel/pinch-style zoom, constrained to a reasonable distance. Clicking a knowledge ball selects it, enlarges it, and populates the right-side inspector. The initial inspector says “Select a knowledge ball” and explains that a glowing node can be clicked to inspect meaning, relationships, and history. Selected nodes show their title, description, “KNOWLEDGE NODE” and ring pills, created/updated date, connection count, related concepts, and integrated state. Use the actual knowledge graph concepts, including Artificial Intelligence, Machine Learning, Programming, Mathematics, Physics, Energy, Sound, Communication, Devices, Security, Ethics, Human Values, Language, Memory, Learning, Time, Vision, Sensors, Robotics, Creativity, plus additional connected concepts such as Curiosity, Planning, Algorithms, Data, Voice, Trust, Emotion, Networks, Adaptation, Questions, Patterns, Control, Imagination, Feedback, Context, Goals, Observation, Attention, Integration, Discovery, Calibration, and Synthesis. Each node can have a short semantic description and related concepts, and display a 2026-09-10 created/updated date.

Overlay a dark translucent glassmorphism HUD over the scene. At the top left show “ULTRON // KNOWLEDGE CORE” with the subtitle “A living interface for memory, learning, and connection.” At the top right show a rounded status badge with a glowing dot and text such as “CALM · ONLINE”; update both dot color and status text as the state changes. Put a lower-left “ACTIVITY STATE” panel containing a two-column grid of buttons: Calm, Listen, Think, Speak, Learn, Excited, Happy, Angry, Sad, and Curious. Calm is initially active with a light fill; clicking any state changes the active button, status label, core/node accent color, waveform energy, pulse intensity, and overall animation energy. Use the visible state palette: calm white, listening cyan, processing violet, learning mint, speaking amber, excited pink, happy green, angry red, sad blue, and curious yellow. Learning should add one new animated knowledge ball from the core into the graph and update the ball count. Listening and speaking should periodically create expanding activity waves; speaking should use browser speech synthesis to say a fixed ULTRON online/status message and then return to calm.

Below the state grid show metrics for “48 knowledge balls” and “72 active links,” followed by a “Talk to ULTRON” microphone button and a “Speak status” button. The microphone control should use browser SpeechRecognition when available: switch to listening, put the recognized transcript into the improvement text area, briefly process it, then respond with browser speech synthesis. Support simple grounded responses: greetings receive a hello/exploration response; identity questions identify ULTRON as a living knowledge interface; mood words switch to the corresponding mood; learning/remember/knowledge/improve requests receive a controlled learning-request response; other text is echoed as something ULTRON heard and is processing. If speech recognition is unavailable, provide a short listening simulation and spoken fallback response.

Put a lower/right dark glass “KNOWLEDGE INSPECTOR” panel. Under the inspector, add an “Improvement Message Box” explaining that the user can propose a controlled change and ULTRON validates it before applying anything. Include a multiline text area with the example placeholder “Example: When I say movie mode, use my authorized entertainment devices.” and an “Analyze proposal” button. On submission, show a result box: empty text yields “NEEDS CLARIFICATION — Describe the improvement you want ULTRON to evaluate.”; proposals involving passwords, disabling security, removing protection, or ignoring permission yield a red “DECLINE” security/authorization message; short or nonspecific proposals yield an amber “NEEDS CLARIFICATION” message asking for a more specific trigger or expected behavior; otherwise show a green “ACCEPT” message saying the proposal can be staged as a controlled configuration change and would require confirmation before applying. Accepted proposals switch the interface to processing; clarification switches to curious.

Keep the visual style minimal, futuristic, calm, and premium: pure black void, white glowing geometry, restrained pale-blue/violet accents, thin low-opacity graph lines, rounded charcoal panels, subtle borders, blur, and high-contrast white typography with uppercase letter spacing. Make the interface responsive: on narrow screens stack/reposition the inspector and activity panels while preserving the central visualization. Include a bottom-center hint reading “DRAG TO ORBIT · SCROLL TO ZOOM · CLICK A KNOWLEDGE BALL.” There are no enemies, weapons, score, lives, or win/lose conditions; the interaction is exploration, state control, voice interaction, node inspection, and proposal validation.