superwhisper · Launch Video Breakdown: Hook, Pacing & Motion Design

Scene-by-scene timeline & spoken transcript

1. 0:00 – 0:04

   The Hook

   Multi-Window Code Environment Reveal

   “(No spoken dialogue — The track features a driving, rhythmic electronic score with a consistent tempo. It incorporates pizzicato synth melodies, a prominent sub-bass line, and 808-style percussion. High-energy audio transitions are present, including a significant riser and sub-bass drop/impact at [0:01].)”

   On screen

   Create a data analysis script that takes the clean data and runs statistical analysis. I'll start by inspecting the cleaned dataset so the script matches its actual schema rather than guessing at column names. Read(data/survey_clean.csv) 1. Read 2,204 lines, 6 columns id, group, age, score_pre, score_post, completed Two numeric outcomes (score_pre, score_post) split across a categorical group (control / treatment). A paired t-test on the pre/post change plus a between-group comparison is the right call. Writing the script now. Write(analysis/statistical_analysis.py) 1. Wrote 31 lines to analysis/statistical_analysis.py import pandas as pd from scipy import stats df = pd.read_csv("data/survey_clean.csv") df = df[df.completed] # drop incomplete responses df["d_score"] = df.score_post - df.score_pre # Paired t-test: did scores move + overall? t, p = stats.ttest_rel(df.score_post, df.score_pre) print(f"paired t = {t:.1f}, p = {p:.3f})") # Between-group: treatment vs control on the change df_group = df[df.group == "treatment"] ct = df[df.group == "control"].delta Update(shape-generator.html) Edit file shape-generator.html <div shape-cell { width: 40px; } Please create a data analysis script that takes the clean data and runs statistical analysis. To help create a statistical analysis script. First, I need to understand what kind of data we're working with and what statistical tests are required. Since no specific data or requirements were provided, I'll create a general-purpose statistical analysis template that can be adapted to different datasets. # statistical_analysis.py # General-purpose statistical analysis import pandas as pd import numpy as np import matplotlib.pyplot as plt import seaborn as sns from scipy import stats from sklearn.linear_model import LinearRegression from statsmodels.formula.api import ols from statsmodels.stats.anova import anova_lm from statsmodels.graphics.gofplots import ProbPlot directory: ~/Development/superwhisper-scripts To get started, describe a task or try one of these commands: /init - create an AGENT.md file with instructions for CodeX /status - show current session configuration /approvals - choose what CodeX can do without approval /models - choose what models and reasoning to use /review - review any changes and find issues Heads up, you have less than 5% of your weekly limit left. Run /status for a breakdown. examine this project and tell me how it works Explored Search rg -files Read ARCHITECTURE.md NOTE_TO_AI.md LORE.md MAIN.js event.js core.js index.html CodeAgent (0.120.0) model: High /model to change directory: ~/Development/superwhisper-scripts Heads up, you have less than 5% of your weekly limit left. Run /status for a breakdown. In shape-generator.js you can add a nice morphing transition when I click generated. Like they morph from square to circle and vice verse or just fade out, but not instant like it is now. Explored Search rg -files Read shape-generator.html I'll add a smooth morphing transition with fade and border-radius animation when generating new shapes. Update(shape-generator.html)

   Camera

   The camera rapidly zooms out from a single terminal window to reveal a complex, multi-window desktop environment, then quickly transitions to black.

   Motion

   Rapid zoom-out, multi-window reveal, quick cut.

2. 0:04 – 0:08

   Product Reveal

   AI Agent Interaction

   “(No spoken dialogue — The electronic score continues with a consistent tempo. An impactful sub-bass hit occurs at [0:08].)”

   On screen

   homerow • neil/homerow for (const metric of metrics) { const current = aggregated.get(metric.userId) ?? 0; aggregated.set(metric.userId, current + metric.sessionDuration); } return aggregated; } export function filterActiveUsers( metrics: UserMetrics[], minActions: number = 5 ): UserMetrics[] { return metrics.filter(m => m.actionsPerformed.length >= minActions); } Dictate or type your response Voice to text Dismiss esc Send Add unit tests for these functions and handle the case where the metrics list is empty?

   Camera

   Static, centered view of a single code editor window with an AI agent interaction prompt.

   Motion

   Text typing animation, UI element reveal (voice to text, dismiss, send buttons).

3. 0:08 – 0:14

   Feature Teaser

   Multi-Agent Workflow & Suggestions

   “(No spoken dialogue — The electronic score maintains its driving rhythm. A sustained synth pad provides an atmospheric layer.)”

   On screen

   homerow • neil/homerow export function filterActiveUsers( metrics: UserMetrics[], minActions: number = 5 ): UserMetrics[] { return metrics.filter(m => m.actionsPerformed.length >= minActions); } Add unit tests for these functions and handle the case where the metrics list is empty? Dictate or type your response Voice to text Dismiss esc Send landing-page • neil/superwhisper-website <button type="submit" disabled={status === 'sending'}> {status === 'sending' ? 'Sending...' : 'Send Message'} </button> </form> } The form posts to /api/contact which doesn't exist yet. Do you want me to: 1. Create the API route with email forwarding ? 2. Just store submissions in your database for now? 3. Add validation with error messages first? Let me know which direction you'd prefer. Dictate or type your response Voice to text Dismiss esc Send Create the API route with email forwarding

   Camera

   The camera rapidly zooms out to reveal multiple code editor windows, then zooms into a new, single window showing an AI agent offering development suggestions.

   Motion

   Rapid zoom-out, multi-window reveal, zoom-in, text typing animation, UI element interaction.

4. 0:14 – 0:17

   Feature Teaser

   Proactive AI Assistance

   “(No spoken dialogue — The electronic score continues with its consistent tempo and atmospheric synth pad.)”

   On screen

   landing-page • neil/superwhisper-website <button {isOpen ? <X size={24} /> : <Menu size={24} />} </button> </div> {isOpen && ( <MobileMenu onClose={() => setIsOpen(false)} /> )} </nav> ); } Dictate or type your response Voice to text Dismiss esc Send I also noticed you don't have any active state on the nav links to show which page the user is on. Want me to add that? I could highlight the current page with an underline or a different color.

   Camera

   Static, centered view of a code editor window with an AI agent proactively suggesting UI/UX improvements.

   Motion

   Text typing animation, UI element reveal.

5. 0:17 – 0:26

   Call to Action

   Integration Announcement & Branding

   “(No spoken dialogue — The electronic score builds slightly in intensity, leading to a final impactful sub-bass drop/impact at the end.)”

   On screen

   Agent integration is now available in Grok Build superwhisper

   Camera

   Static, wide shot of a gradient background. Text fades in, then the product logo fades in.

   Motion

   Text fade-in, logo fade-in, subtle background gradient animation.