@lebtiga/sonic-agent 1.0.0

package/plugin/skills/brand-philosophy/SKILL.md

@@ -0,0 +1,77 @@
+---
+name: brand-philosophy
+version: 1.0.0
+description: Run a brand philosophy discovery session before building a site. Produces a Brand Brief — voice, tone, positioning, target customer, brand pillars. This becomes the source of truth for design and copy decisions downstream. Use this BEFORE any design or content work.
+---
+
+# Brand Philosophy Skill
+
+The discovery step that ALL good agencies do before designing anything. Without this, sites end up generic and forgettable. With it, every downstream decision has a reason.
+
+## When to use this
+
+- At the **start** of any new authority site build
+- Before invoking `visual-identity`
+- Before invoking `frontend-design` or `wordpress-pro` for design work
+- Before generating any copy with `programmatic-seo`
+
+## Output
+
+A `BRAND-BRIEF.md` saved at the project root containing:
+
+1. **Target Customer** — demographic, psychographic, pain points
+2. **Positioning Statement** — why this business vs competitors
+3. **Brand Attitude** — 3-5 adjectives (e.g. "premium, trustworthy, direct")
+4. **Voice & Tone** — how the brand speaks (formal/casual, expert/peer, urgent/calm)
+5. **Three Brand Pillars** — non-negotiable values that show up everywhere
+6. **What We're NOT** — explicit anti-positioning
+7. **The Feeling** — what visitors should feel after 10 seconds on the site
+
+## The Discovery Flow
+
+Sonic asks these questions in order. **Don't skip questions** even if the user seems impatient — each one informs design decisions later.
+
+### Round 1 — The Customer
+
+> Before we design anything, let's get clear on who we're building this for.
+>
+> 1. **Who is the ideal customer for this business?** Be specific — "homeowners" isn't enough. Age range, income level, who they call when they have this problem.
+> 2. **What's the moment they go looking for this service?** Are they panicking (burst pipe), planning (annual maintenance), or comparison shopping (new install)?
+> 3. **What's their biggest fear about hiring someone in this category?** Bad work? Getting overcharged? Not understanding what they need?
+
+### Round 2 — The Positioning
+
+> 4. **Who are 2-3 competitors you'd lose business to?** Real names if you have them.
+> 5. **What does this business do differently or better than those competitors?** Even small things count.
+> 6. **What's a common belief in this industry that you actually disagree with?** (This is gold for differentiation.)
+
+### Round 3 — The Attitude
+
+> 7. **Pick 3 brands you admire (any industry).** Not direct competitors — brands whose VIBE feels right. Why those?
+> 8. **If this brand were a person, what kind of person?** (e.g. "the experienced contractor who explains things plainly" vs "the slick salesperson with the white truck")
+> 9. **What's the ONE thing you absolutely don't want this site to feel like?** (Cheesy? Cheap? Corporate? Generic?)
+
+### Round 4 — The Pillars
+
+> 10. **What 3 values does the business genuinely live by?** Not aspirational — actual operating principles. (e.g. "Show up on time" / "Quote it before you do it" / "If we break it, we fix it free")
+
+### Round 5 — The Feeling
+
+> 11. **A homeowner lands on the site, reads for 10 seconds, then leaves. What do you want them to feel?** (Trust? Urgency? Curiosity? Relief? "This is the one"?)
+
+## After the Q&A
+
+Synthesize the answers into a 1-page Brand Brief. **Be specific and quotable** — generic answers ("we're trustworthy and reliable") get rewritten until they're concrete ("we're the only HVAC company in Tampa that texts you a 30-min ETA window before showing up").
+
+Save to: `BRAND-BRIEF.md` in the project workspace root.
+
+## Then hand off
+
+Once the user approves the Brand Brief, invoke the `visual-identity` skill next.
+
+## What NOT to do
+
+- ❌ Don't generate generic answers if the user gives short replies — push back, ask follow-ups
+- ❌ Don't skip rounds because "we kind of covered it" — every round informs different design decisions
+- ❌ Don't accept "we're like X but better" — that's not a positioning statement
+- ❌ Don't move to design until the brief is signed off

package/plugin/skills/freepik-spaces/SKILL.md

@@ -0,0 +1,122 @@
+# Freepik Spaces Skill
+
+Automate Freepik Spaces workflows via browser control. Create, configure, and run AI-powered visual pipelines for product shots, marketing campaigns, and video content.
+
+## Prerequisites
+
+- Freepik account with Premium/credits
+- OpenClaw Browser Relay extension installed and attached to Chrome
+- Logged into Freepik in Chrome
+
+## Core Concepts
+
+### Node Types
+
+| Node Type | Inputs | Outputs | Purpose |
+|-----------|--------|---------|---------|
+| **Upload** | None | Media | Import local files |
+| **Media** | None | Media | Use Freepik stock/history |
+| **Text** | None | Text | Prompts, lists, notes |
+| **Image Generator** | Text + Reference | Generated Image | AI image creation |
+| **Video Generator** | Text + Start/End Images | Generated Video | AI video creation |
+| **Assistant** | Text + Reference | Text | AI writing/prompts |
+| **Image Upscaler** | Image | Upscaled Image | Enhance resolution |
+| **Camera Angles** | Image | Angle Variations | Generate perspectives |
+| **List** | Text items | Multiple outputs | Fan out to generators |
+| **Container** | Multiple media | Grouped display | Organize outputs |
+
+### Connectors
+
+- **Direction**: Output (top-right) → Input (bottom-left)
+- **One-to-many**: Single output can feed multiple inputs
+- **One-directional**: Data flows source → target only
+
+### Running Workflows
+
+- **Run single node**: Click Run button (bottom-right of node)
+- **Run from here**: Executes node + all downstream nodes
+- **Prerequisites**: All required inputs must be connected before Run activates
+
+## Workflow Patterns
+
+### Product Demo Pipeline
+```
+Product Input → Component List → 3D Renders
+                             → Target Audience → Persona Images
+                             → Locations → Scene Images
+All outputs → Composite Shots → Final Campaign Images/Videos
+```
+
+### SaaS Demo Pipeline
+```
+Product Description → Feature List → Dashboard Screenshots
+                                 → UI Mockups
+                   → User Personas → User in Context Images
+                   → Use Cases → Scenario Visualizations
+All outputs → Product + User + Scenario Composites → Demo Video
+```
+
+## Browser Automation
+
+### Key Selectors
+
+```javascript
+// Navigation
+"button:has-text('New space')"           // Create new space
+"button:has-text('Templates')"           // Templates tab
+"button:has-text('+ Add element')"       // Add node menu
+
+// Node Types (in Add menu)
+"button:has-text('Text')"
+"button:has-text('Image Generator')"
+"button:has-text('Video Generator')"
+"button:has-text('Assistant')"
+"button:has-text('Upscaler')"
+
+// Node Actions
+"button:has-text('Run')"                 // Execute node
+"button:has-text('Run from here')"       // Execute chain
+```
+
+### Common Operations
+
+1. **Create Space**: Click "New space" button
+2. **Add Node**: Click "+ Add element" → select node type
+3. **Edit Text Node**: Click node → type content
+4. **Connect Nodes**: Drag from output point to input point
+5. **Run Workflow**: Click Run on final node or "Run from here"
+
+## Templates
+
+### SaaS Demo Template
+See `templates/saas-demo.json` for the full workflow definition.
+
+**Stages:**
+1. Product Input (name, description, features)
+2. Feature Visualization (dashboard, UI elements)
+3. User Persona Generation
+4. Use Case Scenarios
+5. Product + User Composites
+6. Video Generation
+
+## Usage
+
+```bash
+# Create a new SaaS demo workflow
+# 1. Attach browser relay to Freepik tab
+# 2. Navigate to Spaces
+# 3. Use this skill to automate template creation
+```
+
+## Files
+
+- `SKILL.md` - This documentation
+- `docs/` - Detailed guides and research
+- `templates/` - Workflow template definitions
+- `scripts/` - Automation helper scripts
+
+## Resources
+
+- [Freepik Spaces Docs](https://www.freepik.com/ai/docs/introduction-to-spaces)
+- [Nodes & Connectors](https://www.freepik.com/ai/docs/nodes-and-connectors)
+- [Your First Space](https://www.freepik.com/ai/docs/your-first-space)

package/plugin/skills/freepik-spaces/docs/automation-guide.md

@@ -0,0 +1,233 @@
+# Freepik Spaces Automation Guide
+
+Step-by-step browser automation patterns for building workflows.
+
+## Prerequisites
+
+1. OpenClaw Browser Relay extension installed
+2. Attached to Chrome tab on Freepik
+3. Logged into Freepik account
+
+## Common Actions
+
+### Navigate to Spaces
+
+```javascript
+// Direct URL
+browser.navigate({ targetUrl: "https://www.freepik.com/pikaso/spaces" })
+
+// Or via menu
+browser.act({ kind: "click", ref: "spaces-menu-item" })
+```
+
+### Create New Space
+
+```javascript
+// Click "New space" button
+browser.act({ kind: "click", ref: "new-space-button" })
+// Ref usually: button with text "New space"
+```
+
+### Add a Node
+
+1. Click "+ Add element" (left sidebar, typically first button)
+2. Select node type from menu:
+   - Text
+   - Upload
+   - Media
+   - Image Generator
+   - Video Generator
+   - Assistant
+   - Image Upscaler
+   - Camera Angles
+
+```javascript
+// Open add menu
+browser.act({ kind: "click", ref: "add-element-button" })
+
+// Select node type (look for the type name)
+browser.act({ kind: "click", ref: "text-node-option" })
+```
+
+### Edit Text Node
+
+1. Click on the text node to select it
+2. Start typing (or click into the text area)
+3. Use toolbar for formatting
+
+```javascript
+// Select node
+browser.act({ kind: "click", ref: "text-node" })
+
+// Type content
+browser.act({ kind: "type", ref: "text-node-textarea", text: "Your prompt here" })
+```
+
+### Configure Image Generator
+
+1. Click on Image Generator node
+2. Adjust settings in panel:
+   - Model dropdown
+   - Aspect ratio
+   - Number of images
+
+```javascript
+// Select generator node
+browser.act({ kind: "click", ref: "image-generator-node" })
+
+// Open model dropdown
+browser.act({ kind: "click", ref: "model-dropdown" })
+
+// Select model (e.g., Flux)
+browser.act({ kind: "click", ref: "flux-option" })
+```
+
+### Connect Nodes
+
+1. Hover over source node to reveal output point (top-right)
+2. Click and drag from output to target's input (bottom-left)
+3. Release to create connection
+
+```javascript
+// This requires drag operation
+browser.act({ 
+  kind: "drag", 
+  startRef: "source-node-output", 
+  endRef: "target-node-input" 
+})
+```
+
+### Run a Node
+
+1. Click on the node
+2. Click Run button (bottom-right of node, or in action bar)
+
+```javascript
+// Click Run button on node
+browser.act({ kind: "click", ref: "run-button" })
+
+// Or "Run from here" for chain execution
+browser.act({ kind: "click", ref: "run-from-here" })
+```
+
+### Rename a Node
+
+1. Click node to select
+2. Click node title/name area
+3. Type new name
+
+```javascript
+browser.act({ kind: "click", ref: "node-title" })
+browser.act({ kind: "type", ref: "node-title-input", text: "NEW NAME" })
+browser.act({ kind: "press", key: "Enter" })
+```
+
+## Workflow: Build SaaS Demo Template
+
+### Phase 1: Create Structure
+
+```javascript
+// 1. Create new space
+await browser.navigate({ targetUrl: "https://www.freepik.com/pikaso/spaces" })
+await browser.act({ kind: "click", ref: "new-space-button" })
+
+// 2. Add START text node
+await browser.act({ kind: "click", ref: "add-element" })
+await browser.act({ kind: "click", ref: "text-option" })
+// Type content
+await browser.act({ kind: "type", text: "SaaS Demo Workflow - START HERE" })
+
+// 3. Add Product Input text node
+// Repeat add element flow...
+
+// 4. Add Feature List (List node if available, or multiple Text nodes)
+
+// 5. Add Image Generators for each output type
+// - Dashboard screenshots
+// - UI mockups  
+// - User personas
+// - Scenario images
+
+// 6. Connect all nodes appropriately
+```
+
+### Phase 2: Configure Generators
+
+```javascript
+// For each Image Generator:
+// 1. Click to select
+// 2. Set model to Flux Kontext Pro (good quality, efficient)
+// 3. Set aspect ratio (16:9 for dashboards, 1:1 for portraits)
+// 4. Set num_images to 4
+
+// For prompts, use Assistant node to generate from product description
+```
+
+### Phase 3: Run Workflow
+
+```javascript
+// Run in order:
+// 1. Text/input nodes (just need content)
+// 2. Assistant nodes (to generate prompts)
+// 3. Image generators (parallel if independent)
+// 4. Final composite/hero shots
+```
+
+## Tips for Automation
+
+### Reliable Node Selection
+
+Nodes can be identified by:
+- Their label/title text
+- Generated thumbnails
+- Position on canvas
+- Node type icons
+
+### Handling Loading States
+
+After running a generator, wait for:
+- Loading indicator to disappear
+- Thumbnail preview to appear
+- Run button to re-enable
+
+```javascript
+// Wait pattern
+browser.act({ kind: "wait", textGone: "Generating..." })
+// Or wait for element
+browser.act({ kind: "wait", timeMs: 5000 })
+```
+
+### Error Recovery
+
+If a generation fails:
+1. Click the node
+2. Check input connections
+3. Verify prompt text is valid
+4. Click Run again
+
+### Batch Operations
+
+Use List nodes to fan out:
+1. Create List node with multiple items
+2. Connect to Image Generator
+3. Generator produces output for each list item
+
+## Snapshot Strategy
+
+Take snapshots before/after key operations:
+- After creating new space (capture canvas state)
+- After adding nodes (verify structure)
+- After running generators (capture outputs)
+
+```javascript
+// Capture current state
+const snapshot = await browser.snapshot({ profile: "chrome", targetId: tabId })
+// Parse snapshot to verify expected nodes exist
+```
+
+## Known Limitations
+
+1. **Drag operations**: May be unreliable; prefer click-based alternatives when available
+2. **Dynamic refs**: Node refs change between snapshots; use text/label matching
+3. **Canvas zoom**: May need to zoom out to see all nodes
+4. **Loading times**: AI generation can take 10-60 seconds; build in waits

package/plugin/skills/freepik-spaces/docs/research-notes.md

@@ -0,0 +1,264 @@
+# Freepik Spaces Research Notes
+
+*Compiled: 2026-02-13*
+
+## Overview
+
+Freepik Spaces is a **node-based visual workflow builder** for AI generation. It allows chaining together text prompts, image generators, video generators, and other tools into reusable workflows.
+
+**Key Benefits:**
+- Visual pipeline design on infinite canvas
+- Multiple AI models in one interface
+- Reusable templates
+- Real-time collaboration
+- One subscription covers multiple tools
+
+## Node System Deep Dive
+
+### Input Nodes (Output Only)
+
+**Upload Node**
+- Import files from local computer
+- Supports images, videos
+- No incoming connections
+- Output connects to generators, upscalers, assistant
+
+**Media Node**
+- Select from personal history, Freepik stock, or Community
+- Great for remixing existing content
+- Output connects to generators, upscalers, assistant
+
+**Text Node**
+- Write prompts, instructions, or notes
+- Supports formatting: headings, bold, italic, lists
+- Toolbar appears on click with:
+  - Background color
+  - Rename/duplicate/delete
+  - Preview
+  - Run button
+- Output connects to Image/Video generators, Assistant
+
+### Generator Nodes
+
+**Image Generator**
+- **Inputs**: Text prompt + Reference image (both optional but at least one required)
+- **Output**: Generated images
+- **Settings**:
+  - Model selection (Flux, NanoBanana Pro, etc.)
+  - Aspect ratio
+  - Number of images (1-4)
+- Models available:
+  - Flux Kontext Pro (good consistency, saves credits)
+  - NanoBanana Pro (high quality edits)
+  - And others
+
+**Video Generator**
+- **Inputs**: Text + Start image + End image
+- **Output**: Generated video
+- **Settings**:
+  - Model (Kling, Seedance 4.5, etc.)
+  - Duration
+  - Aspect ratio
+  - Resolution
+  - Number of outputs
+
+**Assistant Node**
+- AI-powered text helper
+- **Inputs**: Text + Reference media
+- **Output**: Generated text
+- Uses:
+  - Generate prompts from descriptions
+  - Describe images/videos
+  - Provide variations/suggestions
+  - Refine ideas
+  
+**Image Upscaler**
+- **Input**: Single image
+- **Output**: Upscaled image
+- **Settings**:
+  - Mode: Creative or Precision
+  - Scale factor
+  - Engine presets
+- Creative mode accepts text prompt for guidance
+
+**Camera Angles**
+- **Input**: Single image
+- **Output**: Same content from different angles
+- Features:
+  - Front/side/top rotations
+  - Camera position control
+  - Start/end frames for video transitions
+- Great for product imagery
+
+### List Node
+
+- Contains multiple text items
+- Each item fans out to connected generators
+- Perfect for batch generation:
+  - Multiple product variations
+  - Multiple personas
+  - Multiple locations
+
+### Container Node
+
+- Groups multiple outputs visually
+- Organizational tool
+- Helps structure complex workflows
+
+## Connectors
+
+### Connection Points
+- **Output**: Top-right corner of node
+- **Input**: Bottom-left corner of node
+- Points visible on hover
+
+### Behavior
+- One-directional (source → target)
+- One output → multiple inputs allowed
+- Click connection line to delete
+
+### Valid Connections by Node Type
+
+| Source Node | Valid Targets |
+|-------------|---------------|
+| Upload/Media | Image Gen, Video Gen, Assistant, Upscaler, Camera Angles |
+| Text | Image Gen, Video Gen, Assistant |
+| Image Generator | Upscaler, Camera Angles, Video Gen, Assistant, Container |
+| Video Generator | Assistant, Container |
+| Assistant | Image Gen, Video Gen, any text-accepting node |
+| Upscaler | Camera Angles, Video Gen, Container |
+| Camera Angles | Upscaler, Video Gen, Container |
+
+## Running Workflows
+
+### Single Node Execution
+1. Click node with selection tool
+2. Click Run button (bottom-right)
+3. Node processes and produces output
+4. Output becomes available for downstream nodes
+
+### Chain Execution
+1. Click Run icon on a node
+2. Select "Run from here"
+3. Executes selected node + all connected downstream nodes
+4. Great for re-running after upstream changes
+
+### Prerequisites
+- All required inputs must be connected
+- Run button stays inactive until prerequisites met
+
+## Best Practices (from community)
+
+### Efficiency
+- Group related nodes visually
+- Use List nodes for batch operations
+- Copy/paste node structures for similar operations
+- Use Assistant to generate prompts rather than writing manually
+
+### Credit Conservation
+- Flux Kontext Pro: Good quality, lower credit cost
+- NanoBanana Pro: Higher quality, higher cost
+- Test with fewer outputs first, then scale up
+
+### Known Issues (Reddit feedback)
+- Assets sometimes mix between projects
+- Saved Spaces occasionally not found
+- Consider phygital plus as alternative for complex node work
+
+## Template Analysis: Dyson Vacuum Workflow
+
+### Structure
+```
+START HERE
+    ↓
+INPUT PRODUCT HERE → PRODUCT BREAKDOWN
+    ↓                        ↓
+TARGET AUDIENCE MAKER → LIST OF COMPONENTS
+    ↓                        ↓
+LIST OF TARGET         3D RENDERS OF
+DEMOGRAPHIC            PRODUCT PARTS
+    ↓                        ↓
+IMAGE GENERATION       360 3D PRODUCT
+OF DEMOGRAPHIC         VIEWS
+    ↓                        ↓
+PRODUCT LOCATION  →   PRODUCT ASSEMBLY
+LIST & GENERATION          ↓
+    ↓                 3D PRODUCT EXPLODED
+    ↓                 DIAGRAM
+    ↓                        ↓
+    └────────┬───────────────┘
+             ↓
+PRODUCT + TARGET DEMOGRAPHIC + LOCATION
+             ↓
+TV COMMERCIAL SHOTS (COMPOSITES)
+             ↓
+TV COMMERCIAL SHOTS W/ EXPLODED COMPONENTS
+```
+
+### Key Patterns
+1. **Parallel branches**: Product visuals + Demographics + Locations
+2. **List nodes**: Fan out to multiple generators
+3. **Convergence**: Multiple streams combine for final output
+4. **Container nodes**: Group related outputs (TV Commercial Shots)
+
+## SaaS Demo Template Design
+
+### Target Output
+- Dashboard/analytics product visuals
+- User personas in context
+- Feature highlights
+- Demo video clips
+
+### Proposed Structure
+```
+PRODUCT INPUT (Name, Description, Key Features)
+    ↓
+    ├── FEATURE LIST → DASHBOARD SCREENSHOTS (Image Gen)
+    │                → UI MOCKUP VARIATIONS
+    │
+    ├── USER PERSONAS → PERSONA LIST → USER IMAGES (Image Gen)
+    │   (Startup founder, Enterprise, etc.)
+    │
+    └── USE CASES → SCENARIO LIST → SCENARIO IMAGES (Image Gen)
+        (Data analysis, Team collaboration, etc.)
+
+All outputs combine:
+    ↓
+PRODUCT + USER + SCENARIO COMPOSITES
+    ↓
+DEMO VIDEO CLIPS (Video Gen) [Phase 2]
+```
+
+### Prompt Templates
+
+**Dashboard Screenshot Prompt:**
+```
+Professional SaaS dashboard UI screenshot, [PRODUCT] analytics platform, 
+clean modern interface, data visualizations, charts and graphs, 
+dark mode / light mode, high-fidelity mockup, web application design
+```
+
+**User Persona Prompt:**
+```
+Professional portrait of [PERSONA TYPE], using laptop showing analytics dashboard,
+modern office environment, natural lighting, business casual attire,
+focused expression, realistic photography style
+```
+
+**Use Case Scenario Prompt:**
+```
+[PERSONA] using [PRODUCT] for [USE CASE], 
+computer screen showing dashboard interface,
+[ENVIRONMENT] setting, professional photography,
+showing [EMOTION/RESULT] from using the product
+```
+
+## Resources
+
+- Official Docs: https://www.freepik.com/ai/docs/introduction-to-spaces
+- Nodes Guide: https://www.freepik.com/ai/docs/nodes-and-connectors
+- First Space Tutorial: https://www.freepik.com/ai/docs/your-first-space
+- YouTube Tutorials:
+  - "Build your first creative AI workflow in Spaces" (Freepik Academy)
+  - "How to get started with Spaces" (Freepik Academy)
+  - "Freepik Just Launched SPACES" (Third-party overview)