Four MIT collections ported — growthOS, openclaw-marketing-skills, Marketing-Skills and pm-skills — as 160 skills, 12 agents and 45 commands, each naming its origin in its own frontmatter so the tree can be checked against THIRD_PARTY_NOTICES rather than trusted. Sixteen names appeared in more than one collection and none of the copies were identical; the longest wins, which lands on the more developed text where they genuinely diverge. A fifth collection is deliberately absent. awesome-growth-hacking-skills carries NO licence, which leaves its author every right by default — and it is a list of links rather than a collection of skills, so there is nothing in it to port and no permission to port it with. It is cited as a directory to read. THE CONNECTOR SKILLS ARE REWRITTEN, not ported. They arrived installing vendor SDKs, running OAuth flows of their own, and ending with long-lived credentials in JSON files under .agents/ and ~/.growthos/ at mode 600. That is a secret on a disk with no rotation path, protected from other users and not from the agent — and one auth implementation per skill. Hanzo holds these connectors in Go, per user, sealed in KMS, so `connect` is the one home for the flow and the platform skills defer to it by provider id: google_ads, meta_ads, x. It is reachable as HTTP or as MCP tools at api.hanzo.ai/v1/mcp — the same routes, typed once in Go, so a tool and a curl cannot disagree. Two had nothing real behind them and now say so. There is no search_console provider, so that skill names what does answer the same questions (google_analytics, google_bigquery) and states that the remedy is a provider in hanzoai/cloud rather than a vendor SDK here. And getxapi-connect was a wrapper around a paid reseller of X's public data with the vendor's key in an env var; it is `listen` now, on our own x connector — a skill named after a vendor makes the vendor the capability and dies with it. The skills referenced paths that only existed inside the source repo, so the shared material they read is ported to reference/ and the paths rewritten. Three skills need the growthOS APPLICATION rather than its data; per the estate's own split an application belongs in hanzoai/ads, social, email or sms as a cloud plugin, so those three say what they need instead of carrying an import that resolves to nothing. Third-party product names are out of our prose and kept in the notices, which is where the licence requires them and the only place they belong. A skill's persona names a role, not somebody else's product. Co-authored-by: Hanzo Dev <[email protected]>
20 KiB
name, description, source
| name | description | source |
|---|---|---|
| showcase-walkthrough | App Walkthrough showcase — step-by-step guided tour with animated cursor, Hotspot pulses, Tooltip explanations, scroll transitions, and ZoomReveal on key moments across multiple screens. Journey-driven from video-profile.yaml. Produces a 30–90s 16:9 MP4 via the AppWalkthrough composition. | melgarafael/growthOS |
Showcase: App Walkthrough
Specialist skill for the App Walkthrough showcase. Activated by the remotion-pro orchestrator when the user selects showcase 8 (walkthrough, tutorial, passo-a-passo, step-by-step, tour). Receives a context object with video_profile, brand_voice, remotion, and briefing.
Overview
The App Walkthrough showcase is the most comprehensive video format — a full guided tour through an entire user journey. An animated cursor navigates the UI step-by-step, Hotspot pulses highlight targets before each click, Tooltip cards explain each action, ScreenScroll transitions move within pages, and ZoomReveal emphasizes key outcomes. Multiple screens are woven into one continuous, smooth flow.
Best for: product onboarding videos, sales demos, tutorial content, user guides, help center videos, investor walkthroughs, support documentation.
Composition ID: AppWalkthrough
Aspect ratio: 16:9 (1920×1080)
FPS: 30
Duration range: 30–90 seconds (900–2700 frames)
Components Used
| Component | File | Purpose in video |
|---|---|---|
BrowserFrame |
components/BrowserFrame.tsx |
Wraps each screen in browser chrome with URL bar |
MobileFrame |
components/MobileFrame.tsx |
Wraps screens in iPhone/Android shell (if device=mobile) |
AnimatedCursor |
components/AnimatedCursor.tsx |
Moves cursor from previous position to next target with spring easing |
ScreenScroll |
components/ScreenScroll.tsx |
Scrolls within a tall screenshot between actions |
ScreenshotOverlay |
components/ScreenshotOverlay.tsx |
Crossfades between different app screens on navigation |
ZoomReveal |
components/ZoomReveal.tsx |
Zooms into result or key UI state after important actions |
AnimatedText |
components/AnimatedText.tsx |
Hook title, step labels, and CTA text |
BackgroundGradient |
components/BackgroundGradient.tsx |
Branded background gradient |
ProgressBar |
components/ProgressBar.tsx |
Accent-colored progress bar at bottom |
BrandWatermark |
components/BrandWatermark.tsx |
Handle overlay at watermark position |
Hotspot (inline component)
Hotspot is rendered as an inline animated div inside the step scene — a pulsing ring that appears at cursor_target coordinates 0.5s before the cursor arrives. Implementation:
// Pulsing ring at target position
// Uses interpolate(frame, [0, 15, 30], [0, 1, 0]) for scale pulse
// Color: brand.colors.accent at 60% opacity
// Size: 40×40px, border-radius: 50%, position: absolute
Tooltip (inline component)
Tooltip is an animated floating label rendered inside the step scene, appearing after the cursor click. Implementation:
// Glassmorphic card (blur: 8px, opacity: 0.85) positioned near cursor_target
// Offsets: if target.y > 540 → tooltip above (y - 80), else below (y + 60)
// Contains: action icon + tooltip text (max 60 chars)
// Entry animation: fade-in + slide-up over 12 frames
// Duration: stays visible for remainder of step scene
Scene Flow Template
Default sequence for a medium-pacing, 3-screen journey with 2 actions per screen:
Scene 1 — Hook (0s–3s, frames 0–89)
AnimatedText: "Here's how [product] works"
Subtext: journey name or product tagline, animation: fade-in delay 15
Scene 2 — Screen 1 Intro (3s–5s, frames 90–149)
BrowserFrame/MobileFrame slides in from bottom
ScreenshotOverlay: first screenshot fades in
AnimatedText: step label appears (e.g., "Step 1: Sign Up")
Scene 3 — Step 1a (5s–9s, frames 150–269)
AnimatedCursor moves from center → target element
Hotspot pulses at target for 15 frames
Click effect fires (ripple)
Tooltip appears: action description
ScreenshotOverlay crossfades to next state
Scene 4 — Step 1b (scroll) (9s–12s, frames 270–359) [if page is scrollable]
ScreenScroll: page scrolls down to next interaction target
Pause at key content for 20 frames
Scene 5 — Transition (12s–14s, frames 360–419)
ScreenshotOverlay crossfades to Screen 2
Step counter increments: "Step 2: [label]"
Scene 6 — Step 2a (14s–18s, frames 420–539)
AnimatedCursor: moves from top-left → target element
Hotspot pulses → click → ripple
Tooltip: action description
ScreenshotOverlay: next state loads
Scene 7 — Zoom Reveal (18s–21s, frames 540–629) [1 zoom per journey]
ZoomReveal on most important result (confirmation screen, success state, key metric)
highlightBox: true, highlightColor: brand.colors.accent
Tooltip: outcome description ("Your dashboard is ready!")
Scene 8 — Screen 3 (21s–27s, frames 630–809) [if 3rd screen exists]
[Repeat cursor + hotspot + click + tooltip pattern]
Scene 9 — CTA (27s–31s, frames 810–929)
Device frame fades out
AnimatedText: CTA headline bounces in
Accent pill: product URL / "Start free trial"
BrandWatermark visible
Pacing multipliers per step type:
fast: cursor travel = 1.5s, tooltip = 1s, scroll = 2smedium: cursor travel = 2s, tooltip = 1.5s, scroll = 3s (default)slow: cursor travel = 3s, tooltip = 2.5s, scroll = 4s
Storyboard Generation Rules
Step 1 — Parse the briefing
product_name → explicit in briefing OR brand-voice.yaml → brand.name
journey_name → explicit in briefing OR first journey in video-profile.yaml
device → "mobile" / "phone" → MobileFrame
→ "browser" / "desktop" / default → BrowserFrame
pacing → explicit in briefing → video-profile.yaml → style.pacing → "medium"
cta_text → explicit in briefing → brand-voice.yaml → cta.default
zoom_on → explicit "zoom on [screen/element]" → auto-select (see Step 4)
Step 2 — Load journey
IF briefing specifies journey name:
Find in video-profile.yaml journeys[] (case-insensitive, fuzzy match)
IF not found: list available journeys, ask user to select
ELSE IF video-profile.yaml has one journey:
Use it automatically
ELSE IF video-profile.yaml has multiple journeys:
Ask: "Which journey should the walkthrough follow? [list names]"
ELSE (no journeys defined):
Ask user to define steps inline (screen path + action + tooltip)
Step 3 — Build WalkthroughStep list
For each screen in the selected journey:
// Map each screen to a WalkthroughStep
{
screenName: screen.name,
screenshot: screen.screenshot,
actions: screen.actions.map(action => mapActionToWalkthroughAction(action))
}
Action string → WalkthroughAction mapping:
| Action string pattern | type | target heuristic | tooltip |
|---|---|---|---|
clica [target] / click [target] |
click |
element center heuristic | "Click [target]" |
clica em [target] |
click |
element center heuristic | "Click on [target]" |
scroll até [section] / scroll to |
scroll |
{ x: 960, y: 540 } | "Scroll to [section]" |
digita [text] / type [text] |
type |
input field heuristic | "Type your [field]" |
preenche [field] / fill [field] |
type |
input field heuristic | "Fill in [field]" |
navega para [page] / navigate to |
click |
nav link heuristic | "Go to [page]" |
hover em [element] / hover [element] |
hover |
element center heuristic | "Hover over [element]" |
submit / confirma / save |
click |
submit button heuristic | "Confirm and continue" |
Target coordinate heuristics (1920×1080, browser frame, content area ~1760×900 offset 80,120):
"Sign Up" / "Get Started" / "CTA button" → { x: 960, y: 500 }
"Login" / "Sign In" → { x: 960, y: 420 }
"Email field" → { x: 760, y: 380 }
"Password field" → { x: 760, y: 460 }
"Submit" / "Continue" / "Next" → { x: 960, y: 600 }
"Navigation menu item" (generic) → { x: 300, y: 160 }
"Search bar" → { x: 760, y: 180 }
"Profile / Avatar" → { x: 1750, y: 160 }
"Settings" → { x: 1700, y: 160 }
"Dashboard / Home card" → { x: 480, y: 520 }
"Sidebar item" → { x: 160, y: 400 }
"Modal close" → { x: 1400, y: 300 }
"Dropdown option" → { x: 760, y: 480 }
Fallback (unknown) → { x: 960, y: 540 }
Step 4 — Select zoom target
The walkthrough should have exactly one ZoomReveal scene. Placement rule:
IF briefing specifies "zoom on [screen/element]":
Use that screen, estimate element coordinates
ELSE:
Auto-select the screen with the most significant outcome:
Priority 1: screen labeled "Dashboard", "Success", "Result", "Confirmation"
Priority 2: last screen in the journey (final state)
Priority 3: screen with action "zoom em" / "zoom on"
ZoomReveal target: { x: 600, y: 300, width: 720, height: 400 } (center of content area)
zoomLevel: 2.0
highlightBox: true
highlightColor: brand.colors.accent
Step 5 — Calculate total duration
hook_duration = 3s
screen_intro_duration = 2s (for first screen only)
per_cursor_travel_s = pacing_cursor[pacing] (1.5 / 2.0 / 3.0)
per_hotspot_s = 0.5 (always)
per_click_effect_s = 0.3 (always)
per_tooltip_s = pacing_tooltip[pacing] (1.0 / 1.5 / 2.5)
per_scroll_action_s = pacing_scroll[pacing] (2.0 / 3.0 / 4.0)
per_type_action_s = max(1.5, text.length × 0.08)
transition_s = 1.5 (between screens)
zoom_reveal_s = 3s (one per video)
cta_duration = 4s
per_step_s = cursor_travel + hotspot + click_effect + tooltip (per action)
+ scroll (if action.type === 'scroll')
+ type_time (if action.type === 'type')
screen_section_s = screen_intro (screen 1 only) + sum(per_step_s for all actions)
total_s = hook + sum(screen_section_s) + transitions × (screens - 1) + zoom + cta
IF total_s < 30: extend tooltip durations by 1.5×, add intro text for each screen
IF total_s > 90: reduce tooltip durations to minimum, merge same-screen actions
Step 6 — Output storyboard YAML
# storyboard.yaml — App Walkthrough
showcase: AppWalkthrough
product: "[product_name]"
journey: "[journey_name]"
device: "[browser|mobile]"
pacing: "[fast|medium|slow]"
total_duration_s: [calculated]
step_count: [total action count across all screens]
scenes:
- id: hook
type: hook
duration_s: 3
text: "Here's how [product_name] works"
subtext: "[journey_name or tagline]"
animation: scale-in
- id: screen-1-intro
type: step
duration_s: 2
step:
screenName: "[screen.name]"
screenshot: "[screen.screenshot]"
actions: []
text: "Step 1: [screen.name]"
animation: slide-up
- id: step-1-1
type: step
duration_s: [calculated]
step:
screenName: "[screen.name]"
screenshot: "[screen.screenshot]"
actions:
- type: "[click|scroll|hover|type]"
target: { x: [N], y: [N] }
tooltip: "[tooltip text]"
duration: [frames]
# next_screenshot for crossfade after click
next_screenshot: "[next state screenshot if available]"
- id: transition-1-2
type: transition
duration_s: 1.5
text: "Step 2: [next screen name]"
animation: crossfade
- id: step-2-1
type: step
duration_s: [calculated]
step:
screenName: "[screen 2 name]"
screenshot: "[screen 2 screenshot]"
actions:
- type: click
target: { x: [N], y: [N] }
tooltip: "[tooltip text]"
duration: [frames]
- id: zoom-reveal
type: step
duration_s: 3
step:
screenName: "[result screen name]"
screenshot: "[result screenshot]"
actions: []
zoom:
target: { x: 600, y: 300, width: 720, height: 400 }
zoomLevel: 2.0
highlightBox: true
text: "[outcome description]"
- id: cta
type: cta
duration_s: 4
text: "[cta headline]"
subtext: "[product URL or 'Start free trial']"
animation: bounce
Composition Mapping
Convert storyboard to Remotion props for the AppWalkthrough composition:
Frame calculation
scene.duration_s × 30 = scene.durationFrames
startFrame[N] = sum of durationFrames[0..N-1]
Props object structure
// AppWalkthroughProps (from types.ts)
{
brand: {
name: "[product_name]",
handle: brand_voice.brand.handle,
colors: {
primary: design_tokens.primary_color,
secondary: design_tokens.secondary_color,
background: "#0F0F1A",
text: "#FFFFFF",
accent: design_tokens.accent_color,
},
watermark_position: "bottom-right",
},
deviceType: "[browser|mobile]",
deviceConfig: {
browser: "chrome",
device: "iphone15",
darkMode: true,
},
showProgressBar: true,
showWatermark: true,
scenes: [
// hook
{
type: "hook",
startFrame: 0,
durationFrames: 90,
text: "Here's how [product] works",
subtext: "[journey name or tagline]",
animation: "scale-in",
},
// step scenes (one per action, or grouped by screen)
{
type: "step",
startFrame: [N],
durationFrames: [N],
text: "Step [N]: [screen name]",
step: {
screenName: "[screen name]",
screenshot: "[path/to/screenshot.png]",
actions: [
{
type: "click",
target: { x: 960, y: 500 },
tooltip: "Click 'Get Started' to begin",
duration: 60, // frames for this action
},
],
},
},
// transition
{
type: "transition",
startFrame: [N],
durationFrames: 45,
text: "Step [N]: [next screen name]",
animation: "wipe",
},
// zoom (one per video)
{
type: "step",
startFrame: [N],
durationFrames: 90,
text: "[outcome description]",
step: {
screenName: "[result screen]",
screenshot: "[result screenshot path]",
actions: [], // no cursor for zoom scene
},
},
// cta
{
type: "cta",
startFrame: [N],
durationFrames: 120,
text: "[cta headline]",
subtext: "[URL or trial text]",
animation: "bounce",
},
],
}
Cursor path interpolation between steps
When building AnimatedCursorProps.path for each action, the cursor's starting position is the previous action's target. For the first action, start from screen center { x: 960, y: 540 }.
path[0] = previous_target (or {x: 960, y: 540} for first action)
path[1] = current action.target with action: "move" (intermediate point if needed)
path[2] = current action.target with action: "click" / "hover"
Add intermediate path points for long diagonal travel (distance > 400px):
midpoint = { x: (start.x + end.x) / 2, y: start.y, action: "move", delay: 5 }
Config Reading
From video-profile.yaml
| Field | Used for |
|---|---|
design_tokens.primary_color |
brand.colors.primary, BrowserFrame bar |
design_tokens.accent_color |
Hotspot color, ProgressBar, CTA pill, ZoomReveal highlight |
design_tokens.heading_font |
step label font-family |
assets.screenshots_dir |
prepended to all screenshot paths |
assets.logo_path |
hook scene overlay if present |
device_frames[] |
validate device choice |
journeys[*] |
source of all screens, screenshots, actions |
style.pacing |
cursor/tooltip/scroll duration multipliers |
style.theme |
background and frame color strategy |
From brand-voice.yaml
| Field | Used for |
|---|---|
brand.name |
product_name fallback, hook text |
brand.handle |
brand.handle in watermark |
brand.tagline |
hook subtext fallback |
cta.default |
CTA text fallback |
automation.autonomy |
output bundle level (manual/semi/auto) |
Briefing overrides (highest priority)
Any explicit value in the user's briefing (journey name, device, pacing, CTA text, zoom target, specific step tooltip text) takes precedence over both YAML files.
Render Command
After generating the composition props, return to the orchestrator with:
composition_id: AppWalkthrough
props: { ...AppWalkthroughProps JSON... }
The orchestrator renders using the discovered Remotion CLI:
cd "${REMOTION_PROJECT}" && npx remotion render src/index.ts AppWalkthrough "${OUTPUT_DIR}/video.mp4" \
--props='{"brand":{...},"scenes":[...],"deviceType":"browser"}' \
--fps=30
For multi-platform (auto autonomy), the orchestrator also runs a 9:16 variant for social:
npx remotion render src/index.ts AppWalkthrough "${OUTPUT_DIR}/video-9x16.mp4" \
--width=1080 --height=1920 --props='...' --fps=30
Briefing Collection UI
When the orchestrator routes here, display this briefing prompt:
APP WALKTHROUGH SHOWCASE
─────────────────────────────────────────────────────────
I'll create a step-by-step guided tour of your product
with cursor navigation, hotspot pulses, and tooltip labels.
Product name:
> [e.g. "" or leave blank to use brand-voice.yaml]
Journey to walk through (from video-profile.yaml):
[list available journey names]
> [select by name or number, or type "custom" to define steps]
Device frame: [browser / mobile] (default: browser)
>
CTA text (optional):
> [e.g. "Start free at .app"]
Pacing: [fast / medium / slow] (default: medium)
>
Zoom focus (optional — which screen to zoom into):
> [e.g. "dashboard" or "confirmation screen" or leave blank for auto]
─────────────────────────────────────────────────────────
If user selects "custom", collect steps inline:
CUSTOM STEPS — Define your walkthrough
Step 1:
Screenshot path: > [e.g. assets/screenshots/home.png]
Screen name: > [e.g. "Landing Page"]
Actions (one per line, blank to finish):
> [e.g. "click Sign Up" or "scroll to pricing"]
> [e.g. "click Get Started"]
>
Add another screen? [y/n] >
Repeat until user types n. Minimum 2 screens required for a walkthrough (otherwise recommend showcase-product-demo instead).
Error Handling
| Error condition | Response |
|---|---|
| Journey not found in video-profile.yaml | List available journeys, offer to define inline |
| Journey has only 1 screen | Warn: "Walkthroughs work best with 2+ screens. Use Product Demo for single-screen videos?" |
| Screenshot file missing | Warn with path, use gray placeholder, list all missing paths at end |
| No actions defined for a screen | Add a default "scroll to bottom" action for that screen |
| Total duration < 25s | Extend tooltip durations, add step-label intro for each screen |
| Total duration > 95s | Ask user which screens to cut, or merge actions into fewer steps |
| More than 8 screens | Warn "8+ screens may make the video too long", suggest splitting into 2 videos |
| Cursor coordinates out of frame | Clamp to (80–1840, 120–1020) content area bounds |
| device not in video-profile.yaml device_frames | Warn, proceed with requested device |
Output
Return to orchestrator:
status: ready
composition_id: AppWalkthrough
storyboard_path: "[OUTPUT_DIR]/storyboard.md"
props: { ...AppWalkthroughProps JSON... }
estimated_duration_s: [N]
screen_count: [N]
step_count: [total actions]
zoom_screen: "[name of zoomed screen]"
The orchestrator handles the render call, thumbnail, captions, and output bundle.