docs: add MCP integration tutorials and update llama.cpp settings for v0.6.6

docs/src/pages/changelog/2025-07-31-llamacpp-tutorials.mdx

@@ -0,0 +1,119 @@
+---
+title: "Jan v0.6.6: Enhanced llama.cpp integration and smarter model management"
+version: 0.6.6
+description: "Major llama.cpp improvements, Hugging Face provider support, and refined MCP experience"
+date: 2025-07-31
+ogImage: "/assets/images/changelog/changelog0.6.6.gif"
+---
+
+import ChangelogHeader from "@/components/Changelog/ChangelogHeader"
+
+<ChangelogHeader title="Jan v0.6.6: Enhanced llama.cpp integration and smarter model management" date="2025-01-31" ogImage="/assets/images/changelog/changelog0.6.6.gif" />
+
+## Highlights 🎉
+
+Jan v0.6.6 delivers significant improvements to the llama.cpp backend, introduces Hugging Face as a
+built-in provider, and brings smarter model management with auto-unload capabilities. This release
+also includes numerous MCP refinements and platform-specific enhancements.
+
+### 🚀 Major llama.cpp Backend Overhaul
+
+We've completely revamped the llama.cpp integration with:
+- **Smart Backend Management**: The backend now auto-updates and persists your settings properly
+- **Device Detection**: Jan automatically detects available GPUs and hardware capabilities
+- **Direct llama.cpp Access**: Models now interface directly with llama.cpp (previously hidden behind Cortex)
+- **Automatic Migration**: Your existing models seamlessly move from Cortex to direct llama.cpp management
+- **Better Error Handling**: Clear error messages when models fail to load, with actionable solutions
+- **Per-Model Overrides**: Configure specific settings for individual models
+
+### 🤗 Hugging Face Cloud Router Integration
+
+Connect to Hugging Face's new cloud inference service:
+- Access pre-configured models running on various providers (Fireworks, Together AI, and more)
+- Hugging Face handles the routing to the best available provider
+- Simplified setup with just your HF token
+- Non-deletable provider status to prevent accidental removal
+- Note: Direct model ID search in Hub remains available as before
+
+### 🧠 Smarter Model Management
+
+New intelligent features to optimize your system resources:
+- **Auto-Unload Old Models**: Automatically free up memory by unloading unused models
+- **Persistent Settings**: Your model capabilities and settings now persist across app restarts
+- **Zero GPU Layers Support**: Set N-GPU Layers to 0 for CPU-only inference
+- **Memory Calculation Improvements**: More accurate memory usage reporting
+
+### 🎯 MCP (Model Context Protocol) Refinements
+
+Enhanced MCP experience with:
+- Tool approval dialog improvements with scrollable parameters
+- Better experimental feature edge case handling
+- Fixed tool call button disappearing issue
+- JSON editing tooltips for easier configuration
+- Auto-focus on "Always Allow" action for smoother workflows
+
+### 📚 New MCP Integration Tutorials
+
+Comprehensive guides for powerful MCP integrations:
+- **Canva MCP**: Create and manage designs through natural language - generate logos, presentations, and marketing materials directly from chat
+- **Browserbase MCP**: Control cloud browsers with AI - automate web tasks, extract data, and monitor sites without complex scripting
+- **Octagon Deep Research MCP**: Access finance-focused research capabilities - analyze markets, investigate companies, and generate investment insights
+
+### 🖥️ Platform-Specific Improvements
+
+**Windows:**
+- Fixed terminal windows popping up during model loading
+- Better process termination handling
+- VCRuntime included in installer for compatibility
+- Improved NSIS installer with app running checks
+
+**Linux:**
+- AppImage now works properly with newest Tauri version and it went from almost 1GB to less than 200MB
+- Better Wayland compatibility
+
+**macOS:**
+- Improved build process and artifact naming
+
+### 🎨 UI/UX Enhancements
+
+Quality of life improvements throughout:
+- Fixed rename thread dialog showing incorrect thread names
+- Assistant instructions now have proper defaults
+- Download progress indicators remain visible when scrolling
+- Better error pages with clearer messaging
+- GPU detection now shows accurate backend information
+- Improved clickable areas for better usability
+
+### 🔧 Developer Experience
+
+Behind the scenes improvements:
+- New automated QA system using CUA (Computer Use Automation)
+- Standardized build process across platforms
+- Enhanced error stream handling and parsing
+- Better proxy support for the new downloader
+- Reasoning format support for advanced models
+
+### 🐛 Bug Fixes
+
+Notable fixes include:
+- Factory reset no longer fails with access denied errors
+- OpenRouter provider stays selected properly
+- Model search in Hub shows latest data only
+- Temporary download files are cleaned up on cancel
+- Legacy threads no longer appear above new threads
+- Fixed encoding issues on various platforms
+
+## Breaking Changes
+
+- Models previously managed by Cortex now interface directly with llama.cpp (automatic migration included)
+- Some sampling parameters have been removed from the llama.cpp extension for consistency
+- Cortex extension is deprecated in favor of direct llama.cpp integration
+
+## Coming Next
+
+We're working on expanding MCP capabilities, improving model download speeds, and adding more provider
+integrations. Stay tuned!
+
+Update your Jan or [download the latest](https://jan.ai/).
+
+For the complete list of changes, see the [GitHub release notes](https://github.com/menloresearch/jan/releases/tag/v0.6.6).

docs/src/pages/docs/llama-cpp.mdx

@@ -59,6 +59,8 @@ These are advanced settings. You typically only need to adjust them if models ar
 | **Engine Version** | Shows which version of llama.cpp you're running | Check compatibility with newer models |
 | **Check Updates** | Downloads newer engine versions | When new models require updated engine |
 | **Backend Selection** | Choose the version optimized for your hardware | After installing new graphics cards or when performance is poor |
+| **Auto Update Engine** | Automatically updates llama.cpp to latest version | Enable for automatic compatibility with new models |
+| **Auto-Unload Old Models** | Unloads unused models to free memory | Enable when running multiple models or low on memory |
 
 ## Hardware Backends
 
@@ -138,8 +140,12 @@ These control how efficiently models run:
 | Setting | What It Does | Recommended Value | Impact |
 |---------|-------------|------------------|---------|
 | **Continuous Batching** | Process multiple requests at once | Enabled | Faster when using multiple tools or having multiple conversations |
-| **Parallel Operations** | How many requests to handle simultaneously | 4 | Higher = more multitasking, but uses more memory |
+| **Threads** | Number of threads for generation | -1 (auto) | -1 uses all logical cores, adjust for specific needs |
-| **CPU Threads** | How many processor cores to use | Auto-detected | More threads can speed up CPU processing |
+| **Threads (Batch)** | Threads for batch and prompt processing | -1 (auto) | Usually same as Threads setting |
+| **Batch Size** | Logical maximum batch size | 2048 | Higher allows more parallel processing |
+| **uBatch Size** | Physical maximum batch size | 512 | Controls memory usage during batching |
+| **GPU Split Mode** | How to distribute model across GPUs | Layer | Layer mode is most common for multi-GPU setups |
+| **Main GPU Index** | Primary GPU for processing | 0 | Change if you want to use a different GPU |
 
 ## Memory Settings
 
@@ -148,15 +154,53 @@ These control how models use your computer's memory:
 | Setting | What It Does | Recommended Value | When to Change |
 |---------|-------------|------------------|----------------|
 | **Flash Attention** | More efficient memory usage | Enabled | Leave enabled unless you have problems |
-| **Caching** | Remember recent conversations | Enabled | Speeds up follow-up questions |
+| **Disable mmap** | Don't memory-map model files | Disabled | Enable if experiencing crashes or pageouts |
-| **KV Cache Type** | Memory precision trade-off | f16 | Change to q8_0 or q4_0 if running out of memory |
+| **MLock** | Keep model in RAM, prevent swapping | Disabled | Enable if you have enough RAM and want consistent performance |
-| **mmap** | Load models more efficiently | Enabled | Helps with large models |
 | **Context Shift** | Handle very long conversations | Disabled | Enable for very long chats or multiple tool calls |
+| **Disable KV Offload** | Keep KV cache on CPU | Disabled | Enable if GPU memory is limited |
+| **KV Cache K Type** | Memory precision for keys | f16 | Change to q8_0 or q4_0 if running out of memory |
+| **KV Cache V Type** | Memory precision for values | f16 | Change to q8_0 or q4_0 if running out of memory |
+| **KV Cache Defragmentation** | Threshold for cache cleanup | 0.1 | Lower values defragment more often |
 
 ### KV Cache Types Explained
-- **f16**: Most stable, uses more memory
+- **f16**: Full 16-bit precision, uses more memory but highest quality
-- **q8_0**: Balanced memory usage and quality
+- **q8_0**: 8-bit quantized, balanced memory usage and quality
-- **q4_0**: Uses least memory, slight quality loss
+- **q4_0**: 4-bit quantized, uses least memory, slight quality loss
+
+## Advanced Settings
+
+These settings are for fine-tuning model behavior and advanced use cases:
+
+### Text Generation Control
+
+| Setting | What It Does | Default Value | When to Change |
+|---------|-------------|---------------|----------------|
+| **Max Tokens to Predict** | Maximum tokens to generate | -1 (infinite) | Set a limit to prevent runaway generation |
+| **Custom Jinja Chat Template** | Override model's chat format | Empty | Only if model needs special formatting |
+
+### RoPE (Rotary Position Embedding) Settings
+
+| Setting | What It Does | Default Value | When to Change |
+|---------|-------------|---------------|----------------|
+| **RoPE Scaling Method** | Context extension method | None | For models that support extended context |
+| **RoPE Scale Factor** | Context scaling multiplier | 1 | Increase for longer contexts |
+| **RoPE Frequency Base** | Base frequency for RoPE | 0 (auto) | Usually loaded from model |
+| **RoPE Frequency Scale Factor** | Frequency scaling factor | 1 | Advanced tuning only |
+
+### Mirostat Sampling
+
+| Setting | What It Does | Default Value | When to Change |
+|---------|-------------|---------------|----------------|
+| **Mirostat Mode** | Alternative sampling method | Disabled | Try V1 or V2 for more consistent output |
+| **Mirostat Learning Rate** | How fast it adapts | 0.1 | Lower for more stable output |
+| **Mirostat Target Entropy** | Target perplexity | 5 | Higher for more variety |
+
+### Output Constraints
+
+| Setting | What It Does | Default Value | When to Change |
+|---------|-------------|---------------|----------------|
+| **Grammar File** | Constrain output format | Empty | For structured output (JSON, code, etc.) |
+| **JSON Schema File** | Enforce JSON structure | Empty | When you need specific JSON formats |
 
 ## Troubleshooting Common Issues
 
@@ -184,23 +228,34 @@ These control how models use your computer's memory:
 
 **For most users:**
 1. Use the default backend that Jan installs
-2. Leave all performance settings at defaults
+2. Enable Auto Update Engine for automatic compatibility
-3. Only adjust if you experience problems
+3. Leave all performance settings at defaults
+4. Only adjust if you experience problems
 
 **If you have an NVIDIA graphics card:**
-1. Download the appropriate CUDA backend
+1. Select the appropriate CUDA backend from the dropdown (e.g., `avx2-cuda-12-0`)
 2. Make sure GPU Layers is set high in model settings
-3. Enable Flash Attention
+3. Keep Flash Attention enabled
+4. Set Main GPU Index if you have multiple GPUs
 
 **If models are too slow:**
-1. Check you're using GPU acceleration
+1. Check you're using GPU acceleration (CUDA/Metal/Vulkan backend)
-2. Try enabling Continuous Batching
+2. Enable Continuous Batching
-3. Close other applications using memory
+3. Increase Batch Size and uBatch Size
+4. Close other applications using memory
 
 **If running out of memory:**
-1. Change KV Cache Type to q8_0
+1. Enable Auto-Unload Old Models
-2. Reduce Context Size in model settings
+2. Change KV Cache K/V Type to q8_0 or q4_0
-3. Try a smaller model
+3. Reduce Context Size in model settings
+4. Enable MLock if you have sufficient RAM
+5. Try a smaller model
+
+**For advanced users:**
+1. Experiment with Mirostat sampling for more consistent outputs
+2. Use Grammar/JSON Schema files for structured generation
+3. Adjust RoPE settings for models with extended context support
+4. Fine-tune thread counts based on your CPU
 
 <Callout type="info">
 Most users can run Jan successfully without changing any of these settings. The defaults are chosen to work well on typical hardware.

docs/src/pages/docs/mcp-examples/browser/browserbase.mdx

@@ -1,6 +1,6 @@
 ---
 title: Browserbase MCP
-description: Automate browsers with AI-powered natural language commands through Browserbase.
+description: Control browsers with natural language through Browserbase's cloud infrastructure.
 keywords:
   [
     Jan,
@@ -19,33 +19,42 @@ import { Callout, Steps } from 'nextra/components'
 
 # Browserbase MCP
 
-[Browserbase MCP](https://docs.browserbase.com/integrations/mcp/introduction) brings browser automation to Jan through natural language commands. Instead of writing complex selectors and automation scripts, you tell the AI what to do and it figures out how to do it.
+[Browserbase MCP](https://docs.browserbase.com/integrations/mcp/introduction) gives AI models actual browser control through cloud infrastructure. Built on Stagehand, it lets you navigate websites, extract data, and interact with web pages using natural language commands.
 
-Built on Stagehand, this integration lets AI models control browsers, navigate websites, fill forms, and extract data using plain English instructions.
+The integration provides real browser sessions that AI can control, enabling tasks that go beyond simple web search APIs.
 
 ## Available Tools
 
-### Core Browser Automation
+<Callout type="info">
-- `browserbase_stagehand_navigate`: Navigate to any URL
+Browserbase's MCP tools evolve over time. This list reflects current capabilities but may change.
-- `browserbase_stagehand_act`: Perform actions using natural language ("click the login button")
+</Callout>
-- `browserbase_stagehand_extract`: Extract text content from pages
+
-- `browserbase_stagehand_observe`: Find and identify page elements
+### Multi-Session Tools
-- `browserbase_screenshot`: Capture page screenshots
+- `multi_browserbase_stagehand_session_create`: Create parallel browser sessions
+- `multi_browserbase_stagehand_session_list`: Track active sessions
+- `multi_browserbase_stagehand_session_close`: Clean up sessions
+- `multi_browserbase_stagehand_navigate_session`: Navigate in specific session
+
+### Core Browser Actions
+- `browserbase_stagehand_navigate`: Navigate to URLs
+- `browserbase_stagehand_act`: Perform actions ("click the login button")
+- `browserbase_stagehand_extract`: Extract text content
+- `browserbase_stagehand_observe`: Find page elements
+- `browserbase_screenshot`: Capture screenshots
 
 ### Session Management
-- `browserbase_session_create`: Create or reuse browser sessions
+- `browserbase_session_create`: Create or reuse sessions
 - `browserbase_session_close`: Close active sessions
-- Multi-session variants for parallel automation workflows
 
 ## Prerequisites
 
 - Jan with MCP enabled
-- Browserbase API key and Project ID from [dashboard.browserbase.com](https://dashboard.browserbase.com)
+- Browserbase account (includes 60 minutes free usage)
-- Model with tool calling support
+- Model with strong tool calling support
 - Node.js installed
 
-<Callout type="info">
+<Callout type="warning">
-Browserbase MCP works best with models that have strong tool calling capabilities. Claude 3.5+ Sonnet, GPT-4o, and similar models perform reliably.
+Currently, only the latest Anthropic models handle multiple tools reliably. Other models may struggle with the full tool set.
 </Callout>
 
 ## Setup
@@ -59,190 +68,208 @@ Browserbase MCP works best with models that have strong tool calling capabilitie
 
 ### Get Browserbase Credentials
 
-1. Visit [dashboard.browserbase.com](https://dashboard.browserbase.com)
+1. Sign up at [browserbase.com](https://browserbase.com)
-2. Create account or sign in
+   - Email verification required
-3. Copy your **API Key** and **Project ID**
+   - Phone number authentication
+   - Thorough security process
 
-[PLACEHOLDER: Screenshot of Browserbase dashboard showing API key and project ID locations]
+2. Access your dashboard and copy:
+   - **API Key** 
+   - **Project ID**
+
+![Browserbase dashboard showing API key and project ID](../../_assets/browserbase.png)
 
 ### Configure MCP Server
 
-Click `+` in MCP Servers section and choose your installation method:
+Click `+` in MCP Servers section:
 
-#### Option 1: Remote URL (Recommended)
+**NPM Package Configuration:**
-
-For the simplest setup with hosted infrastructure:
-
-**Configuration:**
-- **Server Name**: `browserbase`
-- **URL**: Get from [smithery.ai](https://smithery.ai) with your Browserbase credentials
-
-[PLACEHOLDER: Screenshot of smithery.ai configuration page]
-
-#### Option 2: NPM Package
-
-For local installation:
-
-**Configuration:**
 - **Server Name**: `browserbase`
 - **Command**: `npx`
 - **Arguments**: `@browserbasehq/mcp-server-browserbase`
 - **Environment Variables**:
   - Key: `BROWSERBASE_API_KEY`, Value: `your-api-key`
   - Key: `BROWSERBASE_PROJECT_ID`, Value: `your-project-id`
-  - Key: `GEMINI_API_KEY`, Value: `your-gemini-key` (required for Stagehand)
 
-[PLACEHOLDER: Screenshot of Jan MCP server configuration form with Browserbase settings]
+![Jan MCP server configuration with Browserbase settings](../../_assets/browserbase3.png)
 
 ### Verify Setup
 
-Check server status shows as active in the MCP Servers list.
+Check the tools bubble in chat to confirm Browserbase tools are available:
 
-[PLACEHOLDER: Screenshot showing active Browserbase MCP server in Jan]
+![Chat interface showing available Browserbase tools](../../_assets/browserbase2.png)
 
-### Model Configuration
+## Real Usage Example
 
-Use a compatible model with tool calling enabled:
+### Live Information Query
-
-- **Anthropic Claude 3.5+ Sonnet**
-- **OpenAI GPT-4o**
-- **Google Gemini Pro**
-
-[PLACEHOLDER: Screenshot showing model selection with tools enabled]
-
-## Usage
-
-Start a new chat with your tool-enabled model. Browserbase tools will appear in the available tools list.
-
-[PLACEHOLDER: Screenshot showing Browserbase tools in the tools panel]
-
-### Basic Navigation
 
 ```
-Navigate to github.com and take a screenshot
+Which sports matches are happening right now in Australia (irrespective of the sport)?
 ```
 
-The AI will:
+This simple query demonstrates browser automation in action:
-1. Create a browser session
-2. Navigate to the specified URL
-3. Capture a screenshot
-4. Return the image
 
-### Form Interaction
+1. **Tool Activation**
+   - Model creates browser session
+   - Navigates to sports websites
+   - Extracts current match data
 
+![Model using browser tools to search for information](../../_assets/browserbase5.png)
+
+2. **Results Delivery**
+   - Real-time match information
+   - Multiple sports covered
+   - Current scores and timings
+
+![Final response with Australian sports matches](../../_assets/browserbase6.png)
+
+The AI successfully found:
+- AFL matches with live scores
+- NRL games in progress
+- Upcoming Rugby Union fixtures
+
+## Common Issues
+
+### Tool Call Failures
+
+Sometimes tool calls fail due to parsing issues:
+
+![Tool call error showing parsing problem](../../_assets/browserbase7.png)
+
+**Solutions:**
+- Try rephrasing your prompt
+- Disable unnecessary tools
+- Use simpler, more direct requests
+- Switch to Claude 3.5+ Sonnet if using another model
+
+### Model Limitations
+
+Most models struggle with multiple tools. If experiencing issues:
+- Start with single-purpose requests
+- Build complexity gradually
+- Consider which tools are actually needed
+- Expect some trial and error initially
+
+## Usage Limits
+
+**Free Tier:**
+- 60 minutes of browser time included
+- Sessions auto-terminate after 5 minutes inactivity
+- Can adjust timeout in Browserbase dashboard
+- Usage visible in dashboard analytics
+
+**Session Management:**
+- Each browser session counts against time
+- Close sessions when done to conserve minutes
+- Multi-session operations consume time faster
+
+## Practical Use Cases
+
+### Real-Time Data Collection
 ```
-Go to example.com/contact, fill out the contact form with name "John Doe" and email "john@example.com", then submit it
+Check current prices for MacBook Pro M4 at major Australian retailers and create a comparison table.
 ```
 
-The automation will:
+### Form Testing
-- Navigate to the contact page
-- Locate form fields using AI vision
-- Fill in the specified information
-- Submit the form
-
-### Data Extraction
-
 ```
-Visit news.ycombinator.com and extract the titles of the top 10 stories
+Navigate to myservice.gov.au and walk through the Medicare claim process, documenting each required field.
 ```
 
-This will:
+### Content Monitoring
-- Navigate to Hacker News
-- Identify story elements
-- Extract and structure the title data
-- Return as organized text
-
-### Multi-Step Workflows
-
 ```
-Go to GitHub, search for "javascript frameworks", click on the first repository, and tell me about its README content
+Visit ABC News Australia and extract the top 5 breaking news headlines with their timestamps.
 ```
 
-Complex workflows work seamlessly:
+### Multi-Site Analysis
-- Performs the search
-- Navigates to results
-- Extracts repository information
-- Summarizes findings
-
-## Advanced Features
-
-### Multi-Session Management
-
-For parallel browser automation:
-
 ```
-Create two browser sessions: one for monitoring product prices on site A, another for checking inventory on site B
+Compare flight prices from Sydney to Tokyo next week across Qantas, Jetstar, and Virgin Australia.
 ```
 
-Each session maintains independent state, cookies, and context.
+### Automated Verification
+```
+Check if our company is listed correctly on Google Maps, Yelp, and Yellow Pages, noting any discrepancies.
+```
 
-### Custom Configuration
+## Advanced Techniques
 
-The MCP server supports various configuration options:
+### Session Reuse
+```
+Create a browser session, log into LinkedIn, then search for "AI engineers in Melbourne" and extract the first 10 profiles.
+```
 
-- **Proxies**: Enable IP rotation and geo-location testing
+### Parallel Operations
-- **Advanced Stealth**: Bypass detection systems (Scale Plan required)
+```
-- **Custom Viewports**: Set specific browser dimensions
+Create three browser sessions: monitor stock prices on ASX, check crypto on CoinSpot, and track forex on XE simultaneously.
-- **Cookie Injection**: Pre-populate authentication state
+```
 
-### AI Model Selection
+### Sequential Workflows
+```
+Go to seek.com.au, search for "data scientist" jobs in Sydney, apply filters for $150k+, then extract job titles and companies.
+```
 
-Browserbase MCP defaults to Gemini 2.0 Flash but supports multiple AI models:
+## Optimization Tips
-- **Gemini 2.0 Flash** (default, fastest)
-- **GPT-4o** (high accuracy)
-- **Claude 3.5 Sonnet** (excellent reasoning)
 
-## Use Cases
+**Prompt Engineering:**
+- Be specific about what to extract
+- Name exact websites when possible
+- Break complex tasks into steps
+- Specify output format clearly
 
-### E-commerce Monitoring
+**Tool Selection:**
-Track product prices, availability, and competitor information across multiple sites simultaneously.
+- Use multi-session only when needed
+- Close sessions promptly
+- Choose observe before act when possible
+- Screenshot sparingly to save time
 
-### Lead Generation
+**Error Recovery:**
-Extract contact information and business data from directories, social platforms, and company websites.
+- Have fallback prompts ready
-
+- Start simple, add complexity
-### Content Research
+- Watch for timeout warnings
-Gather articles, posts, and media from various sources for analysis and reporting.
+- Monitor usage in dashboard
-
-### Quality Assurance
-Automated testing of web applications, form submissions, and user workflows.
-
-### Market Intelligence
-Monitor competitor activities, pricing changes, and product launches.
-
-### Data Migration
-Extract structured data from legacy systems or poorly documented APIs.
 
 ## Troubleshooting
 
 **Connection Issues:**
-- Verify API key and Project ID accuracy
+- Verify API key and Project ID
-- Check Node.js installation (if using NPM method)
+- Check Browserbase service status
-- Restart Jan application
+- Ensure NPX can download packages
-- Confirm Browserbase account has sufficient credits
+- Restart Jan after configuration
 
-**Tool Calling Problems:**
+**Browser Failures:**
-- Ensure model has tool calling enabled
+- Some sites block automation
-- Try Claude 3.5+ Sonnet or GPT-4o for best results
+- Try different navigation paths
-- Check MCP server shows as active
+- Check if site requires login
+- Verify target site is accessible
 
-**Automation Failures:**
+**Performance Problems:**
-- Use specific, descriptive instructions
+- Reduce concurrent sessions
-- Break complex tasks into smaller steps
+- Simplify extraction requests
-- Check browser console for JavaScript errors
+- Check remaining time quota
-- Verify target website doesn't block automation
+- Consider upgrading plan
 
-**Performance Issues:**
+**Model Struggles:**
-- Use appropriate viewport sizes for your use case
+- Too many tools overwhelm most models
-- Enable proxies only when needed
+- Claude 3.5+ Sonnet most reliable
-- Choose efficient AI models (Gemini Flash for speed)
+- Reduce available tools if needed
-- Close unused browser sessions
+- Use focused, clear instructions
 
-<Callout type="warning">
+<Callout type="info">
-Browserbase has usage limits based on your plan. Monitor session usage to avoid interruptions.
+Browser automation is complex. Expect occasional failures and be prepared to adjust your approach.
 </Callout>
 
+## Browserbase vs Browser Use
+
+| Feature | Browserbase | Browser Use |
+|---------|-------------|-------------|
+| **Infrastructure** | Cloud browsers | Local browser |
+| **Setup Complexity** | API key only | Python environment |
+| **Performance** | Consistent | System dependent |
+| **Cost** | Usage-based | Free (local resources) |
+| **Reliability** | High | Variable |
+| **Privacy** | Cloud-based | Fully local |
+
 ## Next Steps
 
-Browserbase MCP transforms browser automation from a programming task into a conversation. Instead of learning complex automation frameworks, you describe what you want and the AI handles the implementation details.
+Browserbase MCP provides genuine browser automation capabilities, not just web search. This enables complex workflows like form filling, multi-site monitoring, and data extraction that would be impossible with traditional APIs.
 
-The combination of Jan's privacy-focused local processing with Browserbase's cloud browser infrastructure provides the best of both worlds: secure local AI reasoning with scalable remote automation capabilities.
+The cloud infrastructure handles browser complexity while Jan maintains conversational privacy. Just remember: with great browser power comes occasional parsing errors.

docs/src/pages/docs/mcp-examples/browser/browseruse.mdx

@@ -1,269 +0,0 @@
----
-title: Browser Use MCP
-description: Control browsers with natural language through open source Browser Use automation.
-keywords:
-  [
-    Jan,
-    MCP,
-    Model Context Protocol,
-    Browser Use,
-    browser automation,
-    web scraping,
-    open source,
-    headless browser,
-    tool calling,
-  ]
----
-
-import { Callout, Steps } from 'nextra/components'
-
-# Browser Use MCP
-
-[Browser Use MCP](https://docs.browser-use.com/customize/mcp-server) exposes open source browser automation as tools for AI models. Unlike complex automation frameworks that require programming knowledge, Browser Use lets you control browsers through natural language commands.
-
-This MCP server acts as a bridge between Jan and Browser Use's automation capabilities, enabling AI models to navigate websites, interact with elements, and extract content without writing code.
-
-## Available Tools
-
-### Navigation
-- `browser_navigate`: Navigate to URLs or open new tabs
-- `browser_go_back`: Navigate back in browser history
-
-### Interaction
-- `browser_click`: Click elements by index
-- `browser_type`: Type text into input fields
-- `browser_scroll`: Scroll pages up or down
-
-### State & Content
-- `browser_get_state`: Get current page state with interactive elements
-- `browser_extract_content`: AI-powered content extraction from pages
-
-### Tab Management
-- `browser_list_tabs`: List all open browser tabs
-- `browser_switch_tab`: Switch between tabs
-- `browser_close_tab`: Close specific tabs
-
-## Prerequisites
-
-- Jan with MCP enabled
-- Python 3.8+ installed
-- Model with tool calling support
-- Optional: OpenAI API key for content extraction
-
-<Callout type="info">
-Browser Use works with any model that supports tool calling. Claude 3.5+ Sonnet and GPT-4o provide the most reliable results.
-</Callout>
-
-## Setup
-
-### Enable MCP
-
-1. Go to **Settings** > **MCP Servers**
-2. Toggle **Allow All MCP Tool Permission** ON
-
-![MCP settings page with toggle enabled](../../_assets/mcp-on.png)
-
-### Install Browser Use
-
-Browser Use requires Python installation. Install via pip or uv:
-
-```bash
-pip install "browser-use[cli]"
-# or
-uv pip install "browser-use[cli]"
-```
-
-### Configure MCP Server
-
-Click `+` in MCP Servers section:
-
-**Configuration:**
-- **Server Name**: `browser-use`
-- **Command**: `uvx`
-- **Arguments**: `browser-use[cli] --mcp`
-- **Environment Variables**: 
-  - Key: `OPENAI_API_KEY`, Value: `your-openai-key` (optional, for content extraction)
-
-[PLACEHOLDER: Screenshot of Jan MCP server configuration form with Browser Use settings]
-
-### Install Browser Dependencies
-
-Browser Use needs Playwright browsers installed:
-
-```bash
-playwright install chromium
-```
-
-### Verify Setup
-
-Check server status shows as active in the MCP Servers list.
-
-[PLACEHOLDER: Screenshot showing active Browser Use MCP server in Jan]
-
-### Model Configuration
-
-Use a tool-enabled model:
-
-- **Anthropic Claude 3.5+ Sonnet**
-- **OpenAI GPT-4o**
-- **Google Gemini Pro**
-
-[PLACEHOLDER: Screenshot showing model selection with tools enabled]
-
-## Usage
-
-Start a new chat with your tool-enabled model. Browser Use tools will appear in the available tools list.
-
-[PLACEHOLDER: Screenshot showing Browser Use tools in the tools panel]
-
-### Basic Navigation
-
-```
-Navigate to reddit.com and tell me what you see
-```
-
-The AI will:
-1. Launch a browser session
-2. Navigate to Reddit
-3. Capture the page state
-4. Describe the content
-
-### Element Interaction
-
-```
-Go to google.com, search for "browser automation tools", and click the first result
-```
-
-This performs:
-- Navigation to Google
-- Finding the search input
-- Typing the search query
-- Clicking the search button
-- Clicking the first result
-
-### Content Extraction
-
-```
-Visit news.ycombinator.com and extract the top 5 story titles with their URLs
-```
-
-With OpenAI API key configured, this will:
-- Navigate to Hacker News
-- Use AI to identify story elements
-- Extract structured data
-- Return organized results
-
-### Multi-Tab Workflows
-
-```
-Open YouTube and GitHub in separate tabs, then tell me what's trending on both platforms
-```
-
-Browser Use handles:
-- Opening multiple tabs
-- Switching between tabs
-- Extracting content from each
-- Comparing information
-
-## Advanced Features
-
-### State Inspection
-
-The `browser_get_state` tool provides detailed page information:
-- Current URL and title
-- All interactive elements with indices
-- Tab information
-- Optional screenshots
-
-### Smart Element Selection
-
-Browser Use uses AI-powered element detection. Instead of complex CSS selectors, elements are identified by:
-- Index numbers from page state
-- Natural language descriptions
-- Visual context understanding
-
-### Persistent Sessions
-
-Browser sessions remain active between commands, enabling:
-- Multi-step workflows
-- Authentication persistence
-- Complex navigation sequences
-
-## Use Cases
-
-### Web Research
-Automate information gathering from multiple sources, compare data across sites, and compile research reports.
-
-### E-commerce Monitoring
-Track product availability, price changes, and competitor analysis across shopping platforms.
-
-### Social Media Management
-Monitor mentions, extract engagement metrics, and analyze content performance across platforms.
-
-### Quality Assurance Testing
-Automated testing of web applications, form submissions, and user journey validation.
-
-### Content Aggregation
-Collect articles, posts, and media from various sources for analysis and curation.
-
-### Lead Generation
-Extract contact information and business intelligence from directories and professional networks.
-
-## Configuration Options
-
-### Downloads Directory
-Files download to `~/Downloads/browser-use-mcp/` by default.
-
-### Action Timing
-Default wait time between actions is 0.5 seconds, preventing rate limiting and ensuring page loads.
-
-### Session Persistence
-Browser sessions stay alive between commands, maintaining login states and navigation history.
-
-## Troubleshooting
-
-**Installation Issues:**
-- Verify Python 3.8+ is installed
-- Install Browser Use with `pip install "browser-use[cli]"`
-- Run `playwright install chromium` for browser dependencies
-
-**Server Connection Problems:**
-- Check MCP server shows as active in Jan
-- Restart Jan after configuration changes
-- Verify command path: `uvx browser-use[cli] --mcp`
-
-**Browser Launch Failures:**
-- Ensure Playwright browsers are installed
-- Check system permissions for browser launching
-- Try running `playwright install --help` for troubleshooting
-
-**Tool Calling Issues:**
-- Confirm model has tool calling enabled
-- Use Claude 3.5+ Sonnet or GPT-4o for best results
-- Check that Browser Use tools appear in the tools panel
-
-**Content Extraction Not Working:**
-- Add OpenAI API key to environment variables
-- Verify API key has sufficient credits
-- Check API key permissions for text processing
-
-<Callout type="warning">
-Browser Use launches actual browser instances. Monitor system resources with multiple concurrent sessions.
-</Callout>
-
-## Browser Use vs Browserbase
-
-| Feature | Browser Use | Browserbase |
-|---------|-------------|-------------|
-| **Infrastructure** | Local browser instances | Cloud-hosted browsers |
-| **Cost** | Free (open source) | Usage-based pricing |
-| **Setup** | Python installation required | API key only |
-| **Performance** | Local system dependent | Optimized cloud infrastructure |
-| **Privacy** | Full local control | Data processed in cloud |
-| **Scalability** | Limited by local resources | Highly scalable |
-
-## Next Steps
-
-Browser Use MCP brings powerful browser automation to Jan without requiring cloud dependencies or usage fees. The open source approach gives you complete control over the automation environment while maintaining Jan's privacy-focused philosophy.
-
-For scenarios requiring high-scale automation or specialized browser features, consider Browserbase MCP as a complementary cloud-based solution.

docs/src/pages/docs/mcp-examples/data-analysis/jupyter.mdx

@@ -1,347 +0,0 @@
----
-title: Jupyter MCP
-description: Real-time Jupyter notebook interaction and code execution through MCP integration.
-keywords:
-  [
-    Jan,
-    MCP,
-    Model Context Protocol,
-    Jupyter,
-    data analysis,
-    code execution,
-    notebooks,
-    Python,
-    visualization,
-    tool calling,
-  ]
----
-
-import { Callout, Steps } from 'nextra/components'
-
-# Jupyter MCP
-
-[Jupyter MCP Server](https://github.com/datalayer/jupyter-mcp-server) enables real-time interaction with Jupyter notebooks, allowing AI models to edit, execute, and document code for data analysis and visualization. Instead of just generating code suggestions, AI can actually run Python code and see the results.
-
-This integration transforms Jan from a code-suggesting assistant into a fully capable data science partner that can execute analysis, create visualizations, and iterate based on actual results.
-
-## Available Tools
-
-### Core Notebook Operations
-- `insert_execute_code_cell`: Add and run code cells with immediate execution
-- `append_markdown_cell`: Add documentation and explanations
-- `get_notebook_info`: Retrieve notebook structure and metadata
-- `read_cell`: Examine existing cell content and outputs
-- `modify_cell`: Edit existing cells and re-execute
-
-### Advanced Features
-- **Real-time synchronization**: See changes as they happen
-- **Smart execution**: Automatic retry and adjustment when cells fail
-- **Output feedback**: AI learns from execution results to improve code
-- **Multi-cell workflows**: Complex analysis across multiple cells
-
-## Prerequisites
-
-- Jan with MCP enabled
-- Python 3.8+ with uv package manager
-- Docker installed
-- Model with tool calling support
-- Basic understanding of Jupyter notebooks
-
-<Callout type="info">
-This setup requires running JupyterLab locally. The MCP server connects to your local Jupyter instance for real-time interaction.
-</Callout>
-
-## Setup
-
-### Enable MCP
-
-1. Go to **Settings** > **MCP Servers**
-2. Toggle **Allow All MCP Tool Permission** ON
-
-![MCP settings page with toggle enabled](../../_assets/mcp-on.png)
-
-### Install uv Package Manager
-
-If you don't have uv installed:
-
-```bash
-# macOS and Linux
-curl -LsSf https://astral.sh/uv/install.sh | sh
-
-# Windows
-powershell -c "irm https://astral.sh/uv/install.ps1 | iex"
-```
-
-### Create Python Environment
-
-Create an isolated environment for Jupyter:
-
-```bash
-# Create and activate environment
-uv venv jupyter-mcp
-source jupyter-mcp/bin/activate  # Linux/macOS
-# or
-jupyter-mcp\Scripts\activate     # Windows
-
-# Install required packages
-uv pip install jupyterlab==4.4.1 jupyter-collaboration==4.0.2 ipykernel
-
-# Handle dependency conflicts
-uv pip uninstall pycrdt datalayer_pycrdt
-uv pip install datalayer_pycrdt==0.12.17
-```
-
-### Start JupyterLab Server
-
-Launch JupyterLab with the required configuration:
-
-```bash
-jupyter lab --port 8888 --IdentityProvider.token MY_TOKEN --ip 0.0.0.0
-```
-
-This starts JupyterLab accessible at `http://localhost:8888` with token `MY_TOKEN`.
-
-[PLACEHOLDER: Screenshot of JupyterLab running with token authentication]
-
-### Configure MCP Server
-
-Click `+` in MCP Servers section and choose your OS configuration:
-
-#### macOS and Windows Configuration
-
-**Configuration:**
-- **Server Name**: `jupyter`
-- **Command**: `docker`
-- **Arguments**: 
-  ```
-  run -i --rm -e ROOM_URL -e ROOM_TOKEN -e ROOM_ID -e RUNTIME_URL -e RUNTIME_TOKEN datalayer/jupyter-mcp-server:latest
-  ```
-- **Environment Variables**: 
-  - Key: `ROOM_URL`, Value: `http://host.docker.internal:8888`
-  - Key: `ROOM_TOKEN`, Value: `MY_TOKEN`
-  - Key: `ROOM_ID`, Value: `notebook.ipynb`
-  - Key: `RUNTIME_URL`, Value: `http://host.docker.internal:8888`
-  - Key: `RUNTIME_TOKEN`, Value: `MY_TOKEN`
-
-#### Linux Configuration
-
-**Configuration:**
-- **Server Name**: `jupyter`
-- **Command**: `docker`
-- **Arguments**: 
-  ```
-  run -i --rm -e ROOM_URL -e ROOM_TOKEN -e ROOM_ID -e RUNTIME_URL -e RUNTIME_TOKEN --network=host datalayer/jupyter-mcp-server:latest
-  ```
-- **Environment Variables**: 
-  - Key: `ROOM_URL`, Value: `http://localhost:8888`
-  - Key: `ROOM_TOKEN`, Value: `MY_TOKEN`
-  - Key: `ROOM_ID`, Value: `notebook.ipynb`
-  - Key: `RUNTIME_URL`, Value: `http://localhost:8888`
-  - Key: `RUNTIME_TOKEN`, Value: `MY_TOKEN`
-
-[PLACEHOLDER: Screenshot of Jan MCP server configuration with Jupyter settings]
-
-### Create Target Notebook
-
-In JupyterLab, create a new notebook named `notebook.ipynb` (matching your `ROOM_ID`).
-
-[PLACEHOLDER: Screenshot of creating new notebook in JupyterLab]
-
-### Verify Setup
-
-Check server status shows as active in the MCP Servers list.
-
-[PLACEHOLDER: Screenshot showing active Jupyter MCP server in Jan]
-
-### Model Configuration
-
-Use a tool-enabled model:
-
-- **Anthropic Claude 3.5+ Sonnet**
-- **OpenAI GPT-4o**
-- **Google Gemini Pro**
-
-[PLACEHOLDER: Screenshot showing model selection with tools enabled]
-
-## Usage
-
-Start a new chat with your tool-enabled model. Jupyter tools will appear in the available tools list.
-
-[PLACEHOLDER: Screenshot showing Jupyter tools in the tools panel]
-
-### Data Analysis Workflow
-
-```
-Load the iris dataset and create a scatter plot showing the relationship between sepal length and petal length, colored by species.
-```
-
-The AI will:
-1. Insert a code cell to load the dataset
-2. Execute the code and verify it works
-3. Create visualization code
-4. Run the plotting code
-5. Display results and iterate if needed
-
-### Exploratory Data Analysis
-
-```
-I have sales data in a CSV file. Load it, show me the first few rows, then create a summary of sales by month with a trend chart.
-```
-
-This produces:
-- Data loading and validation
-- Initial data exploration
-- Monthly aggregation analysis
-- Trend visualization
-- Summary insights
-
-### Machine Learning Pipeline
-
-```
-Build a simple classification model on this dataset. Split the data, train a model, evaluate performance, and show feature importance.
-```
-
-The workflow includes:
-- Data preprocessing
-- Train/test split
-- Model training
-- Performance evaluation
-- Feature importance visualization
-- Results interpretation
-
-### Real-Time Iteration
-
-```
-The plot looks good but the colors are hard to distinguish. Make them more vibrant and add a legend.
-```
-
-The AI will:
-- Identify the relevant plotting cell
-- Modify color scheme
-- Add legend configuration
-- Re-execute the cell
-- Show updated visualization
-
-## Advanced Features
-
-### Multi-Cell Workflows
-
-The MCP server manages complex analyses across multiple cells:
-- Imports and setup in initial cells
-- Data processing in subsequent cells
-- Visualization and analysis in final cells
-- Automatic dependency tracking
-
-### Error Handling and Recovery
-
-When code execution fails:
-- AI receives error messages
-- Automatic troubleshooting and fixes
-- Re-execution with corrections
-- Learning from failure patterns
-
-### Real-Time Collaboration
-
-Changes made directly in JupyterLab are immediately visible to the AI:
-- Bidirectional synchronization
-- Conflict resolution
-- Version tracking
-- Collaborative editing
-
-## Configuration Details
-
-### Environment Variables Explained
-
-- **ROOM_URL**: JupyterLab server URL for notebook access
-- **ROOM_TOKEN**: Authentication token for JupyterLab
-- **ROOM_ID**: Path to target notebook (relative to JupyterLab root)
-- **RUNTIME_URL**: Jupyter kernel server URL for code execution
-- **RUNTIME_TOKEN**: Authentication token for kernel access
-
-### Network Configuration
-
-- **macOS/Windows**: Uses `host.docker.internal` for Docker-to-host communication
-- **Linux**: Uses `--network=host` for direct network access
-- **Port matching**: Ensure MCP configuration matches JupyterLab port
-
-## Use Cases
-
-### Data Science Research
-Interactive analysis, hypothesis testing, and visualization creation with real-time code execution and iteration.
-
-### Educational Tutorials
-Create step-by-step analysis tutorials with executed examples and explanations for learning data science concepts.
-
-### Business Analytics
-Generate reports, dashboards, and insights from business data with automated analysis and visualization.
-
-### Prototype Development
-Rapid prototyping of data processing pipelines, machine learning models, and analytical workflows.
-
-### Code Documentation
-Automatically document analysis processes with markdown cells explaining methodology and results.
-
-### Collaborative Analysis
-Work with AI to explore datasets, test hypotheses, and develop analytical solutions interactively.
-
-## Troubleshooting
-
-**JupyterLab Connection Issues:**
-- Verify JupyterLab is running on the specified port
-- Check token authentication is working
-- Confirm notebook file exists at specified path
-- Test JupyterLab access in browser
-
-**Docker Container Problems:**
-- Ensure Docker is running and accessible
-- Check network configuration for your OS
-- Verify environment variables are set correctly
-- Test Docker container can reach JupyterLab
-
-**Python Environment Issues:**
-- Activate the correct uv environment
-- Install missing packages with `uv pip install`
-- Resolve dependency conflicts
-- Check Python and package versions
-
-**Code Execution Failures:**
-- Verify kernel is running in JupyterLab
-- Check for missing Python packages
-- Examine error messages in notebook
-- Restart Jupyter kernel if needed
-
-**MCP Server Connection:**
-- Check server shows as active in Jan
-- Verify all environment variables are set
-- Restart Jan after configuration changes  
-- Test Docker container manually
-
-<Callout type="warning">
-Jupyter MCP requires both JupyterLab server and Docker to be running. Monitor system resources with active notebook sessions.
-</Callout>
-
-## Security Considerations
-
-**Code Execution:**
-- AI has full Python execution capabilities
-- Review generated code before execution
-- Use isolated Python environments
-- Monitor system resource usage
-
-**Network Access:**
-- JupyterLab server is network accessible
-- Use strong authentication tokens
-- Consider firewall restrictions
-- Monitor access logs
-
-**Data Privacy:**
-- Notebook content is processed by AI models
-- Keep sensitive data in secure environments
-- Review data handling policies
-- Use local-only configurations when needed
-
-## Next Steps
-
-Jupyter MCP transforms Jan into a fully capable data science partner that can execute real Python code, create visualizations, and iterate on analysis based on actual results. This moves beyond code suggestions to genuine collaborative data analysis.
-
-The real-time interaction enables a natural workflow where you describe what you want to analyze, and the AI builds the analysis step-by-step, adjusting based on intermediate results and your feedback.

docs/src/pages/docs/mcp-examples/deepresearch/octagon.mdx

@@ -1,6 +1,6 @@
 ---
 title: Octagon Deep Research MCP
-description: Comprehensive AI-powered research capabilities for technical teams and complex analysis.
+description: Finance-focused deep research with AI-powered analysis through Octagon's MCP integration.
 keywords:
   [
     Jan,
@@ -8,10 +8,10 @@ keywords:
     Model Context Protocol,
     Octagon,
     deep research,
+    financial research,
+    private equity,
+    market analysis,
     technical research,
-    debugging,
-    framework comparison,
-    API research,
     tool calling,
   ]
 ---
@@ -20,33 +20,42 @@ import { Callout, Steps } from 'nextra/components'
 
 # Octagon Deep Research MCP
 
-[Octagon Deep Research MCP](https://docs.octagonagents.com/guide/deep-research-mcp.html) provides enterprise-grade research capabilities designed for technical teams. Instead of spending hours researching frameworks, debugging complex issues, or evaluating tools, you get comprehensive analysis that goes straight from research to working applications.
+[Octagon Deep Research MCP](https://docs.octagonagents.com/guide/deep-research-mcp.html) provides specialized AI research capabilities with a strong focus on financial markets and business intelligence. Unlike general research tools, Octagon excels at complex financial analysis, market dynamics, and investment research.
 
-Claims to be faster than ChatGPT Deep Research, more thorough than Perplexity, and without the rate limits that throttle your workflow.
+The integration delivers comprehensive reports that combine multiple data sources, cross-verification, and actionable insights - particularly useful for understanding market structures, investment strategies, and business models.
 
-## Available Research Capabilities
+## Available Tools
 
-### Technical Research
+### octagon-agent
-- **Complex Debugging**: Root cause analysis across distributed systems, memory leak detection
+Orchestrates comprehensive market intelligence research, particularly strong in:
-- **Framework Comparison**: Performance benchmarking, scalability analysis, ecosystem evaluation
+- Financial market analysis
-- **Package Discovery**: Dependency compatibility, security assessment, bundle size impact
+- Private equity and M&A research
-- **API Research**: Documentation analysis, rate limiting comparison, integration complexity
+- Corporate structure investigations
-- **Performance Optimization**: Code profiling, database tuning, caching strategies
+- Investment strategy evaluation
 
-### Business Intelligence
+### octagon-scraper-agent
-- **Market Research**: Competitive analysis, industry trends, customer behavior
+Specialized web scraping for public and private market data:
-- **Tool Discovery**: Development workflow optimization, CI/CD evaluation
+- SEC filings and regulatory documents
-- **Design Analysis**: UI/UX patterns, accessibility compliance, conversion optimization
+- Company financials and metrics
+- Market transaction data
+- Industry reports and analysis
+
+### octagon-deep-research-agent
+Comprehensive research synthesis combining:
+- Multi-source data aggregation
+- Cross-verification of claims
+- Historical trend analysis
+- Actionable insights generation
 
 ## Prerequisites
 
 - Jan with MCP enabled
-- Octagon account and API key from [octagonagents.com](https://octagonagents.com)
+- Octagon account (includes 2-week Pro trial)
 - Model with tool calling support
 - Node.js installed
 
 <Callout type="info">
-Octagon offers unlimited research runs without rate limits, unlike ChatGPT Pro's 125-task monthly restriction.
+Octagon offers a 2-week Pro trial upon signup, providing full access to their financial research capabilities.
 </Callout>
 
 ## Setup
@@ -60,204 +69,192 @@ Octagon offers unlimited research runs without rate limits, unlike ChatGPT Pro's
 
 ### Get Octagon API Key
 
-1. Sign up at [octagonagents.com](https://octagonagents.com)
+1. Sign up at [Octagon signup page](https://app.octagonai.co/signup/?redirectToAfterSignup=https://app.octagonai.co/api-keys)
-2. Navigate to **API Keys** from the left menu
+2. Navigate to the API playground
-3. Generate a new API key
+3. Copy your API key from the dashboard
-4. Save the key
 
-[PLACEHOLDER: Screenshot of Octagon dashboard showing API key generation]
+![Octagon API playground showing API key location](../../_assets/octagon2.png)
 
 ### Configure MCP Server
 
-Click `+` in MCP Servers section and choose your method:
+Click `+` in MCP Servers section:
 
-#### Option 1: Remote Server (Recommended)
+**NPM Package Configuration:**
-
+- **Server Name**: `octagon-mcp-server`
-**Configuration:**
-- **Server Name**: `octagon-deep-research`
-- **URL**: `https://mcp.octagonagents.com/deep-research/mcp`
-- **Environment Variables**: Leave empty (OAuth authentication)
-
-#### Option 2: NPM Package
-
-**Configuration:**
-- **Server Name**: `octagon-deep-research`
 - **Command**: `npx`
-- **Arguments**: `-y octagon-deep-research-mcp@latest`
+- **Arguments**: `-y octagon-mcp@latest`
 - **Environment Variables**: 
   - Key: `OCTAGON_API_KEY`, Value: `your-api-key`
 
-[PLACEHOLDER: Screenshot of Jan MCP server configuration with Octagon settings]
+![Jan MCP server configuration with Octagon settings](../../_assets/octagon3.png)
 
 ### Verify Setup
 
-Check server status shows as active in the MCP Servers list.
+Check the tools bubble in chat to confirm Octagon tools are available:
 
-[PLACEHOLDER: Screenshot showing active Octagon MCP server in Jan]
+![Chat interface showing available Octagon tools with moonshotai/kimi-k2 model](../../_assets/octagon4.png)
 
-### Model Configuration
+## Real-World Example: Private Equity Analysis
 
-Use a tool-enabled model:
+Here's an actual deep research query demonstrating Octagon's financial analysis capabilities:
 
-- **Anthropic Claude 3.5+ Sonnet**
+### The Prompt
-- **OpenAI GPT-4o**
-- **Google Gemini Pro**
-
-[PLACEHOLDER: Screenshot showing model selection with tools enabled]
-
-## Usage
-
-Start a new chat with your tool-enabled model. Octagon research tools will appear in the available tools list.
-
-[PLACEHOLDER: Screenshot showing Octagon tools in the tools panel]
-
-### Technical Debugging
 
 ```
-Research the latest techniques for spotting memory leaks in large React + Node.js projects. Then build a command-line analyzer that scans a codebase and suggests fixes.
+Break apart the private equity paradox: How did an industry that promises to "unlock value" become synonymous with gutting companies, yet still attracts the world's smartest money? 
+
+Start with the mechanics—how PE firms use other people's money to buy companies with borrowed cash, then charge fees for the privilege. Trace the evolution from corporate raiders of the 1980s to today's trillion-dollar titans like Blackstone, KKR, and Apollo. Use SEC filings, M&A databases, and bankruptcy records to map their empires.
+
+Dig into specific deals that illustrate the dual nature: companies genuinely transformed versus those stripped and flipped. Compare Toys "R" Us's death to Hilton's resurrection. Examine how PE-owned companies fare during economic downturns—do they really have "patient capital" or do they bleed portfolio companies dry through dividend recaps?
+
+Investigate the fee structure that makes partners billionaires regardless of performance. Calculate the real returns after the 2-and-20 (or worse) fee structures. Why do pension funds and endowments keep pouring money in despite academic studies showing they'd do better in index funds?
+
+Explore the revolving door between PE, government, and central banks. How many Fed officials and Treasury secretaries came from or went to PE? Map the political donations and lobbying expenditures that keep carried interest taxed as capital gains.
+
+Address the human cost through labor statistics and case studies—what happens to employees when PE takes over? But also examine when PE genuinely saves failing companies and preserves jobs.
+
+Write this as if explaining to a skeptical but curious friend over drinks—clear language, no jargon without explanation, and enough dry humor to make the absurdities apparent. Think Michael Lewis meets Matt Levine. Keep it under 3,000 words but pack it with hard data and real examples. The goal: help readers understand why PE is simultaneously capitalism's most sophisticated expression and its most primitive.
 ```
 
-The research will:
+![Prompt entered in Jan UI](../../_assets/octagon5.png)
-1. Analyze current memory leak detection methods
+
-2. Compare available tools and libraries
+### Research Process
-3. Identify best practices and patterns
+
-4. Generate working code implementation
+The AI engages multiple Octagon tools to gather comprehensive data:
-5. Provide optimization recommendations
+
+![Kimi model using Octagon tools for research](../../_assets/octagon6.png)
+
+### The Results
+
+Octagon delivers a detailed analysis covering:
+
+**Part 1: The Mechanics Explained**
+![First part of the research report](../../_assets/octagon7.png)
+
+**Part 2: Historical Analysis and Case Studies**
+![Second part showing PE evolution and specific deals](../../_assets/octagon8.png)
+
+**Part 3: Financial Engineering and Human Impact**
+![Final section on fee structures and consequences](../../_assets/octagon9.png)
+
+The report demonstrates Octagon's ability to:
+- Access and analyze SEC filings
+- Compare multiple deal outcomes
+- Calculate real returns after fees
+- Track political connections
+- Assess human impact with data
+
+## Finance-Focused Use Cases
+
+### Investment Research
+```
+Analyze Tesla's vertical integration strategy vs traditional automakers. Include supply chain dependencies, margin analysis, and capital efficiency metrics from the last 5 years.
+```
+
+### Market Structure Analysis
+```
+Map the concentration of market makers in US equities. Who controls order flow, what are their profit margins, and how has this changed since zero-commission trading?
+```
+
+### Corporate Governance
+```
+Investigate executive compensation at the 10 largest US banks post-2008. Compare pay ratios, stock buybacks vs R&D spending, and correlation with shareholder returns.
+```
+
+### Private Market Intelligence
+```
+Track Series B+ funding rounds in AI/ML companies in 2024. Identify valuation trends, investor concentration, and compare to public market multiples.
+```
+
+### Regulatory Analysis
+```
+Examine how Basel III implementation differs across major markets. Which banks gained competitive advantages and why?
+```
+
+### M&A Strategy
+```
+Analyze Microsoft's acquisition strategy under Nadella. Calculate actual vs projected synergies, integration success rates, and impact on market position.
+```
+
+## Technical Research Capabilities
+
+While finance-focused, Octagon also handles technical research:
 
 ### Framework Evaluation
-
 ```
-Compare Next.js, Remix, and Astro for high-traffic e-commerce. Benchmark build speed, runtime performance, and SEO. Build a functional storefront with the winner.
+Compare Kubernetes alternatives for edge computing. Consider resource usage, latency, reliability, and operational complexity with real deployment data.
 ```
 
-This produces:
+### API Economics
-- Detailed performance comparisons
-- Real-world benchmarking data
-- SEO and accessibility analysis
-- Working storefront implementation
-- Deployment recommendations
-
-### API Analysis
-
 ```
-Evaluate leading video-streaming APIs on cost, latency, uptime, and documentation quality. Create a working demo with the top choice.
+Analyze the unit economics of major AI API providers. Include pricing history, usage patterns, and margin estimates based on reported compute costs.
 ```
 
-Results include:
+### Open Source Sustainability
-- Comprehensive API comparison matrix
-- Cost-benefit analysis
-- Integration complexity assessment
-- Working implementation demo
-- Production deployment guide
-
-### Package Discovery
-
 ```
-Compare Python libraries for real-time chat systems. Build a runnable chat server with the best stack, including Docker setup.
+Research funding models for critical open source infrastructure. Which projects are at risk and what are the economic incentives misalignments?
 ```
 
-Delivers:
+## Research Quality
-- Library performance benchmarks
-- Feature comparison analysis
-- Security and maintenance evaluation
-- Complete working implementation
-- Production-ready configuration
 
-## Advanced Research Capabilities
+Octagon's reports typically include:
-
+- **Primary Sources**: SEC filings, earnings calls, regulatory documents
-### Multi-Source Analysis
+- **Quantitative Analysis**: Financial metrics, ratios, trend analysis
-
+- **Comparative Studies**: Peer benchmarking, historical context
-Octagon pulls from extensive technical sources:
+- **Narrative Clarity**: Complex topics explained accessibly
-- Official documentation and changelogs
+- **Actionable Insights**: Not just data, but implications
-- GitHub repositories and issue trackers
-- Stack Overflow and developer forums
-- Performance benchmarking sites
-- Security vulnerability databases
-
-### Cross-Verification
-
-Research findings are cross-checked across multiple sources for accuracy and currency.
-
-### Implementation Focus
-
-Unlike generic research tools, Octagon prioritizes actionable results that lead directly to working code and implementations.
-
-## Use Cases
-
-### Development Teams
-Accelerate technical decision-making with comprehensive framework comparisons, tool evaluations, and architecture research.
-
-### DevOps Engineering
-Research deployment strategies, monitoring solutions, and infrastructure optimization techniques with working examples.
-
-### Product Development
-Analyze market opportunities, competitive landscapes, and user experience patterns to inform product strategy.
-
-### Technical Architecture
-Evaluate technology stacks, performance optimization strategies, and scalability solutions for complex systems.
-
-### Startup Research
-Rapid market analysis, competitive intelligence, and technical feasibility assessments for new ventures.
-
-### Enterprise Migration
-Research migration strategies, compatibility assessments, and risk analysis for large-scale technology transitions.
-
-## Performance Claims
-
-**Speed**: 8-10x faster than ChatGPT Deep Research
-**Depth**: 3x more source coverage than competitors
-**Limits**: No rate restrictions (vs ChatGPT Pro's 125 monthly tasks)
-**Accuracy**: Cross-verified technical details and metrics
 
 ## Troubleshooting
 
 **Authentication Issues:**
-- Verify Octagon account is active
+- Verify API key from Octagon dashboard
-- Check API key format and permissions
+- Check trial status hasn't expired
-- Re-authenticate OAuth connection
+- Ensure correct API key format
-- Confirm account has sufficient credits
+- Contact Octagon support if needed
 
-**Server Connection Problems:**
+**Research Failures:**
-- Check MCP server shows as active in Jan
+- Some queries may exceed scope (try narrowing)
-- Verify internet connectivity to Octagon services
+- Financial data may have access restrictions
-- Restart Jan after configuration changes
+- Break complex queries into parts
-- Test API key validity
+- Allow time for comprehensive research
 
-**Research Quality Issues:**
+**Tool Calling Problems:**
-- Be specific in research requests
+- Not all models handle multiple tools well
-- Break complex queries into focused topics
+- Kimi-k2 via OpenRouter works reliably
-- Specify desired output format (code, analysis, comparison)
+- Claude 3.5+ Sonnet also recommended
-- Include context about your use case
+- Enable tool calling in model settings
 
-**Performance Problems:**
+**Performance Considerations:**
-- Monitor Octagon account usage limits
+- Deep research takes time (be patient)
-- Check network latency to research servers
+- Complex financial analysis may take minutes
-- Try breaking large requests into smaller chunks
+- Monitor API usage in dashboard
-- Verify model has sufficient context window
+- Consider query complexity vs urgency
-
-**Tool Calling Issues:**
-- Ensure model supports tool calling
-- Check that Octagon tools appear in tools panel
-- Try Claude 3.5+ Sonnet for best results
-- Verify MCP permissions are enabled
 
 <Callout type="warning">
-Octagon Deep Research requires an active internet connection and may have usage limits based on your account type.
+Octagon specializes in financial and business research. While capable of technical analysis, it's optimized for market intelligence and investment research.
 </Callout>
 
-## Octagon vs Competitors
+## Pricing After Trial
 
-| Feature | Octagon | ChatGPT Deep Research | Perplexity | Grok DeepSearch |
+After the 2-week Pro trial:
-|---------|---------|----------------------|------------|-----------------|
+- Check current pricing at octagonagents.com
-| **Speed** | 8-10x faster | Baseline | Moderate | Moderate |
+- Usage-based pricing for API access
-| **Rate Limits** | None | 125/month | Limited | Limited |
+- Different tiers for research depth
-| **Source Coverage** | 3x more | Standard | Standard | Standard |
+- Educational discounts may be available
-| **Technical Focus** | Specialized | General | General | General |
+
-| **Code Generation** | Integrated | Separate | Limited | Limited |
+## Octagon vs Other Research Tools
-| **Verification** | Cross-checked | Basic | Basic | Basic |
+
+| Feature | Octagon | ChatGPT Deep Research | Perplexity |
+|---------|---------|----------------------|------------|
+| **Finance Focus** | Specialized | General | General |
+| **Data Sources** | Financial databases | Web-wide | Web-wide |
+| **SEC Integration** | Native | Limited | Limited |
+| **Market Data** | Comprehensive | Basic | Basic |
+| **Research Depth** | Very Deep | Deep | Moderate |
+| **Speed** | Moderate | Slow | Fast |
 
 ## Next Steps
 
-Octagon Deep Research MCP transforms research from a time-consuming manual process into an automated capability that delivers actionable results. Instead of spending hours evaluating options, you get comprehensive analysis with working implementations.
+Octagon Deep Research MCP excels at complex financial analysis that would typically require a team of analysts. The integration provides institutional-quality research capabilities within Jan's conversational interface.
 
-The unlimited research capability removes the bottleneck of monthly limits, enabling development teams to make informed decisions without workflow interruption.
+Whether analyzing market structures, evaluating investments, or understanding business models, Octagon delivers the depth and accuracy that financial professionals expect, while maintaining readability for broader audiences.

docs/src/pages/docs/mcp-examples/design/canva.mdx

@@ -1,6 +1,6 @@
 ---
 title: Canva MCP
-description: Create and edit designs through natural language commands with Canva's official MCP server.
+description: Create and manage designs through natural language commands with Canva's official MCP server.
 keywords:
   [
     Jan,
@@ -19,18 +19,47 @@ import { Callout, Steps } from 'nextra/components'
 
 # Canva MCP
 
-[Canva MCP](https://www.canva.com/newsroom/news/deep-research-integration-mcp-server/) brings professional design capabilities directly into Jan through natural language commands. As the first design platform to offer native MCP integration, Canva lets AI models create presentations, resize graphics, and edit templates without leaving the chat.
+[Canva MCP](https://www.canva.com/newsroom/news/deep-research-integration-mcp-server/) gives AI models the ability to create, search, and manage designs directly within Canva. As the first design platform with native MCP integration, it lets you generate presentations, logos, and marketing materials through conversation rather than clicking through design interfaces.
 
-Instead of switching between apps to create designs, you describe what you need and the AI handles the design work using your Canva account and assets.
+The integration provides comprehensive design capabilities without leaving your chat, though actual editing still happens in Canva's interface.
 
 ## Available Tools
 
-The Canva MCP server provides comprehensive design automation:
+<Callout type="info">
-- **Template Generation**: Create presentations, social media posts, and documents
+Canva's MCP tools may change over time as the integration evolves. This list reflects current capabilities.
-- **Asset Management**: Import, export, and organize design assets  
+</Callout>
-- **Design Editing**: Resize graphics, modify layouts, and update content
+
-- **Content Integration**: Use existing Canva assets and brand elements
+### Design Operations
-- **Format Conversion**: Export designs in multiple formats (PDF, PNG, etc.)
+- **generate-design**: Create new designs using AI prompts
+- **search-designs**: Search docs, presentations, videos, whiteboards
+- **get-design**: Get detailed information about a Canva design
+- **get-design-pages**: List pages in multi-page designs
+- **get-design-content**: Extract content from designs
+- **resize-design**: Adapt designs to different dimensions
+- **get-design-resize-status**: Check resize operation status
+- **get-design-generation-job**: Track AI generation progress
+
+### Import/Export
+- **import-design-from-url**: Import files from URLs as new designs
+- **get-design-import-from-url**: Check import status
+- **export-design**: Export designs in various formats
+- **get-export-formats**: List available export options
+- **get-design-export-status**: Track export progress
+
+### Organization
+- **create-folder**: Create folders in Canva
+- **move-item-to-folder**: Organize designs and assets
+- **list-folder-items**: Browse folder contents
+
+### Collaboration
+- **comment-on-design**: Add comments to designs
+- **list-comments**: View design comments
+- **list-replies**: See comment threads
+- **reply-to-comment**: Respond to feedback
+
+### Legacy Tools
+- **search**: ChatGPT connector (use search-designs instead)
+- **fetch**: Content retrieval for ChatGPT
 
 ## Prerequisites
 
@@ -40,10 +69,6 @@ The Canva MCP server provides comprehensive design automation:
 - Node.js installed
 - Internet connection for Canva API access
 
-<Callout type="info">
-Canva MCP works with both free and paid Canva accounts. Paid accounts have access to premium templates and features.
-</Callout>
-
 ## Setup
 
 ### Enable MCP
@@ -63,19 +88,31 @@ Click `+` in MCP Servers section:
 - **Arguments**: `-y mcp-remote@latest https://mcp.canva.com/mcp`
 - **Environment Variables**: Leave empty (authentication handled via OAuth)
 
-[PLACEHOLDER: Screenshot of Jan MCP server configuration form with Canva settings]
+![Canva MCP server configuration in Jan](../../_assets/canva.png)
 
-### Verify Setup
+### Authentication Process
 
-Check server status shows as active in the MCP Servers list.
+When you first use Canva tools:
 
-[PLACEHOLDER: Screenshot showing active Canva MCP server in Jan]
+1. **Browser Opens Automatically**
+   - Canva authentication page appears in your default browser
+   - Log in with your Canva account
 
-### Authentication
+![Canva authentication page](../../_assets/canva2.png)
 
-The first time you use Canva tools, you'll be prompted to authenticate with your Canva account through OAuth. This grants secure access to your designs and templates.
+2. **Team Selection & Permissions**
+   - Select your team (if you have multiple)
+   - Review permissions the AI will have
+   - Click **Allow** to grant access
 
-[PLACEHOLDER: Screenshot of Canva OAuth authentication flow]
+![Canva team selection and permissions](../../_assets/canva3.png)
+
+The permissions include:
+- Reading your profile and designs
+- Creating new designs
+- Managing folders and content
+- Accessing team brand templates
+- Commenting on designs
 
 ### Model Configuration
 
@@ -85,189 +122,160 @@ Use a tool-enabled model:
 - **OpenAI GPT-4o**
 - **Google Gemini Pro**
 
-[PLACEHOLDER: Screenshot showing model selection with tools enabled]
+## Real-World Usage Example
 
-## Usage
+Here's an actual workflow creating a company logo:
 
-Start a new chat with your tool-enabled model. Canva tools will appear in the available tools list.
+### Initial Setup Confirmation
-
-[PLACEHOLDER: Screenshot showing Canva tools in the tools panel]
-
-### Create Presentations
 
 ```
-Create a 5-slide pitch deck about sustainable energy solutions. Use a professional template with blue and green colors.
+Are you able to access my projects?
 ```
 
-The AI will:
+The AI explains available capabilities:
-1. Access Canva's template library
-2. Select appropriate presentation template
-3. Generate content for each slide
-4. Apply consistent styling
-5. Return the completed presentation
 
-### Design Social Media Content
+![AI response about available actions](../../_assets/canva4.png)
+
+### Design Creation Request
 
 ```
-Create an Instagram post announcing our new product launch. Make it eye-catching with our brand colors.
+Create new designs with AI. Call it "VibeBusiness" and have it be a company focused on superintelligence for the benefit of humanity.
 ```
 
-This generates:
+The AI initiates design generation:
-- Platform-optimized dimensions
-- Brand-consistent styling
-- Engaging visual elements
-- Ready-to-publish format
 
-### Resize and Adapt Designs
+![AI generating design with tool call visible](../../_assets/canva5.png)
 
+### Design Options
+
+The AI creates multiple logo variations:
+
+**First Option:**
+![First logo design option](../../_assets/canva6.png)
+
+**Selected Design:**
+![Selected logo design](../../_assets/canva7.png)
+
+### Final Result
+
+After selection, the AI confirms:
+
+![Final response with design ready](../../_assets/canva8.png)
+
+Clicking the design link opens it directly in Canva:
+
+![Design opened in Canva browser tab](../../_assets/canva9.png)
+
+## Practical Use Cases
+
+### Marketing Campaign Development
 ```
-Take my existing LinkedIn post design and resize it for Twitter, Facebook, and Instagram Stories.
+Create a social media campaign for our new product launch. Generate Instagram posts, Facebook covers, and LinkedIn banners with consistent branding.
 ```
 
-Canva MCP will:
+### Presentation Automation
-- Access your existing design
-- Create versions for each platform
-- Maintain visual consistency
-- Optimize for platform specifications
-
-### Generate Marketing Materials
-
 ```
-Create a flyer for our upcoming workshop on January 15th. Include registration details and make it professional but approachable.
+Search for our Q4 sales presentation and create a simplified 5-slide version for the board meeting.
 ```
 
-Results in:
+### Brand Asset Management
-- Professional layout and typography
+```
-- Clear information hierarchy
+List all designs in our "2025 Marketing" folder and export the approved ones as PDFs.
-- Contact and registration details
+```
-- Print-ready format
 
-## Advanced Features
+### Design Iteration
+```
+Find our company logo designs from last month and resize them for business cards, letterheads, and email signatures.
+```
 
-### Brand Integration
+### Content Extraction
+```
+Extract all text from our employee handbook presentation so I can update it in our documentation.
+```
 
-Canva MCP automatically applies your brand guidelines:
+### Collaborative Review
-- Consistent color schemes
+```
-- Brand fonts and styling
+Add a comment to the new website mockup asking the design team about the color scheme choices.
-- Logo placement
+```
-- Template preferences
 
-### Asset Library Access
+## Workflow Tips
 
-Access your entire Canva library:
+### Effective Design Generation
-- Previously created designs
+- **Be specific**: "Create a minimalist tech company logo with blue and silver colors"
-- Uploaded images and graphics
+- **Specify format**: "Generate an Instagram story template for product announcements"
-- Brand kit elements
+- **Include context**: "Design a professional LinkedIn banner for a AI research company"
-- Purchased premium content
+- **Request variations**: Ask for multiple options to choose from
 
-### Collaborative Workflows
+### Organization Best Practices
+- Create folders before generating multiple designs
+- Use descriptive names for easy searching later
+- Move designs to appropriate folders immediately
+- Export important designs for backup
 
-Designs created through MCP integrate with Canva's collaboration features:
+### Integration Patterns
-- Share with team members
+- Generate designs → Review options → Select preferred → Open in Canva for fine-tuning
-- Collect feedback and comments
+- Search existing designs → Extract content → Generate new versions
-- Track version history
+- Create templates → Resize for multiple platforms → Export all variants
-- Manage permissions
 
-### Export Options
+## Limitations and Considerations
 
-Multiple format support:
+**Design Editing**: While the MCP can create and manage designs, actual editing requires opening Canva's interface.
-- **PDF**: Print-ready documents
-- **PNG/JPG**: Web and social media
-- **MP4**: Video presentations
-- **GIF**: Animated graphics
 
-## Use Cases
+**Project Access**: The integration may not access all historical projects immediately, focusing on designs created or modified after connection.
 
-### Marketing Teams
+**Generation Time**: AI design generation takes a few moments. The tool provides job IDs to track progress.
-Create campaign materials, social media content, and promotional designs with consistent branding across all channels.
 
-### Sales Presentations
+**Team Permissions**: Access depends on your Canva team settings and subscription level.
-Generate pitch decks, proposal documents, and client presentations that reflect current data and messaging.
-
-### Educational Content
-Design lesson materials, infographics, and student handouts that engage and inform effectively.
-
-### Event Planning
-Create invitations, programs, signage, and promotional materials for events and conferences.
-
-### Small Business Marketing
-Develop professional marketing materials without requiring design expertise or expensive software.
-
-### Content Creation
-Generate visual content for blogs, newsletters, and digital marketing campaigns.
-
-## Design Best Practices
-
-### Effective Prompts
-- **Be specific**: "Create a minimalist LinkedIn banner with our logo and tagline"
-- **Include dimensions**: "Design a square Instagram post"
-- **Specify style**: "Use our brand colors and modern typography"
-- **Mention purpose**: "Create a presentation slide explaining our pricing model"
-
-### Brand Consistency
-- Upload brand assets to Canva before using MCP
-- Specify brand elements in prompts
-- Use consistent messaging and visual style
-- Review designs for brand compliance
-
-### Template Selection
-- Choose templates appropriate for your audience
-- Consider platform requirements and limitations
-- Balance creativity with readability
-- Test designs across different devices
 
 ## Troubleshooting
 
 **Authentication Issues:**
-- Clear browser cache and cookies
+- Clear browser cookies for Canva
-- Re-authenticate with Canva account
+- Try logging out and back into Canva
-- Check Canva account permissions
+- Ensure pop-ups aren't blocked for OAuth flow
-- Verify internet connection stability
+- Check team admin permissions if applicable
 
-**Server Connection Problems:**
+**Design Generation Failures:**
-- Confirm MCP server shows as active
+- Verify you have creation rights in selected team
-- Check Node.js installation and version
+- Check Canva subscription limits
-- Restart Jan after configuration changes
+- Try simpler design prompts first
-- Test network connectivity to Canva services
+- Ensure stable internet connection
 
-**Design Creation Failures:**
+**Tool Availability:**
-- Be more specific in design requests
+- Some tools require specific Canva plans
-- Check Canva account limits and quotas
+- Team features need appropriate permissions
-- Verify template availability for your account type
+- Verify MCP server is showing as active
-- Try simpler design requests first
+- Restart Jan after authentication
 
-**Export Issues:**
+**Search Problems:**
-- Check available export formats for your account
+- Use search-designs (not the legacy search tool)
-- Verify file size limitations
+- Be specific with design types and names
-- Ensure sufficient storage space
+- Check folder permissions for team content
-- Try exporting individual elements separately
+- Allow time for new designs to index
-
-**Tool Calling Problems:**
-- Ensure model supports tool calling
-- Check that Canva tools appear in tools panel
-- Try Claude 3.5+ Sonnet or GPT-4o
-- Verify MCP permissions are enabled
 
 <Callout type="warning">
-Canva MCP requires an active internet connection and may have usage limits based on your Canva account type.
+Design generation uses Canva's AI capabilities and may be subject to usage limits based on your account type.
 </Callout>
 
-## Privacy and Security
+## Advanced Workflows
 
-**Data Handling:**
+### Batch Operations
-- Designs remain in your Canva account
+```
-- No content is shared without explicit permission
+Create 5 variations of our product announcement banner, then resize all of them for Twitter, LinkedIn, and Facebook.
-- OAuth provides secure, revocable access
+```
-- All API communications are encrypted
 
-**Account Security:**
+### Content Migration
-- Use strong Canva account passwords
+```
-- Enable two-factor authentication
+Import all designs from [URLs], organize them into a "2025 Campaign" folder, and add review comments for the team.
-- Review connected applications periodically
+```
-- Monitor account activity for unusual access
+
+### Automated Reporting
+```
+Search for all presentation designs created this month, extract their content, and summarize the key themes.
+```
 
 ## Next Steps
 
-Canva MCP transforms design from a separate workflow step into part of your natural conversation with AI. Instead of describing what you want and then creating it separately, you can generate professional designs directly within your AI interactions.
+Canva MCP bridges the gap between conversational AI and visual design. Instead of describing what you want and then manually creating it, you can generate professional designs directly through natural language commands.
 
-This integration positions design as a thinking tool rather than just an output format, enabling more creative and efficient workflows across marketing, sales, education, and content creation.
+The real power emerges when combining multiple tools - searching existing assets, generating new variations, organizing content, and collaborating with teams, all within a single conversation flow.

docs/src/pages/docs/mcp-examples/search/jina.mdx

@@ -1,288 +0,0 @@
----
-title: Jina MCP
-description: Access Jina AI's web search and content extraction APIs through community MCP servers.
-keywords:
-  [
-    Jan,
-    MCP,
-    Model Context Protocol,
-    Jina AI,
-    web search,
-    content extraction,
-    web scraping,
-    fact checking,
-    tool calling,
-  ]
----
-
-import { Callout, Steps } from 'nextra/components'
-
-# Jina MCP
-
-Jina AI provides powerful web search and content extraction APIs, but doesn't offer an official MCP server. The community has built two MCP implementations that bring Jina's capabilities to Jan.
-
-These community servers enable AI models to search the web, extract content from web pages, and perform fact-checking using Jina's infrastructure.
-
-<Callout type="warning">
-These are community-maintained packages, not official Jina AI releases. Exercise standard caution when installing third-party packages and verify their source code before use.
-</Callout>
-
-## Available Servers
-
-### PsychArch/jina-mcp-tools
-Basic implementation with core functionality:
-- `jina_reader`: Extract content from web pages
-- `jina_search`: Search the web
-
-### JoeBuildsStuff/mcp-jina-ai  
-More comprehensive implementation with additional features:
-- `read_webpage`: Advanced content extraction with multiple formats
-- `search_web`: Web search with configurable options
-- `fact_check`: Fact-checking and grounding capabilities
-
-## Prerequisites
-
-- Jan with MCP enabled
-- Jina AI API key from [jina.ai](https://jina.ai/?sui=apikey) (optional for basic features)
-- Model with tool calling support
-- Node.js installed
-
-<Callout type="info">
-Some features work without an API key, but registration provides enhanced capabilities and higher rate limits.
-</Callout>
-
-## Setup
-
-### Enable MCP
-
-1. Go to **Settings** > **MCP Servers**
-2. Toggle **Allow All MCP Tool Permission** ON
-
-![MCP settings page with toggle enabled](../../_assets/mcp-on.png)
-
-### Get Jina API Key
-
-1. Visit [jina.ai](https://jina.ai/?sui=apikey)
-2. Create account or sign in
-3. Generate API key
-4. Save the key
-
-[PLACEHOLDER: Screenshot of Jina AI dashboard showing API key generation]
-
-### Choose Your Server
-
-Select one of the community implementations:
-
-#### Option 1: Basic Server (PsychArch)
-
-**Configuration:**
-- **Server Name**: `jina-basic`
-- **Command**: `npx`
-- **Arguments**: `jina-mcp-tools`
-- **Environment Variables**: 
-  - Key: `JINA_API_KEY`, Value: `your-api-key` (optional)
-
-#### Option 2: Advanced Server (JoeBuildsStuff)
-
-**Configuration:**
-- **Server Name**: `jina-advanced`
-- **Command**: `npx`
-- **Arguments**: `-y jina-ai-mcp-server`
-- **Environment Variables**: 
-  - Key: `JINA_API_KEY`, Value: `your-api-key`
-
-[PLACEHOLDER: Screenshot of Jan MCP server configuration with Jina settings]
-
-### Verify Setup
-
-Check server status shows as active in the MCP Servers list.
-
-[PLACEHOLDER: Screenshot showing active Jina MCP server in Jan]
-
-### Model Configuration
-
-Use a tool-enabled model:
-
-- **Anthropic Claude 3.5+ Sonnet**
-- **OpenAI GPT-4o**
-- **Google Gemini Pro**
-
-[PLACEHOLDER: Screenshot showing model selection with tools enabled]
-
-## Usage
-
-Start a new chat with your tool-enabled model. Jina tools will appear in the available tools list.
-
-[PLACEHOLDER: Screenshot showing Jina tools in the tools panel]
-
-### Web Content Extraction
-
-**Basic Server:**
-```
-Extract the main content from https://example.com/article
-```
-
-**Advanced Server:**
-```
-Read the webpage at https://news.example.com and format it as clean markdown
-```
-
-The advanced server offers multiple extraction modes:
-- **Standard**: Balanced speed and quality
-- **Comprehensive**: Maximum data extraction
-- **Clean Content**: Remove ads and navigation
-
-### Web Search
-
-**Basic Server:**
-```
-Search for "latest developments in quantum computing" and return 5 results
-```
-
-**Advanced Server:**
-```
-Search for machine learning tutorials, limit to 10 results from github.com
-```
-
-The advanced server supports:
-- Configurable result counts
-- Site-specific filtering
-- Multiple output formats
-
-### Fact Checking (Advanced Only)
-
-```
-Fact-check this statement: "The Great Wall of China is visible from space"
-```
-
-Returns:
-- Factuality scores
-- Supporting evidence
-- Reference sources
-- Contradictory information
-
-### GitHub Integration
-
-Both servers handle GitHub URLs intelligently:
-
-```
-Extract the code from https://github.com/owner/repo/blob/main/script.js
-```
-
-GitHub file URLs are automatically converted to raw content for direct access.
-
-## Advanced Features
-
-### Content Formats (Advanced Server)
-
-- **Default**: Jina's native markdown format
-- **Markdown**: Structured with headers and links
-- **Text**: Plain text only
-- **HTML**: Raw HTML content
-- **Screenshot**: Visual page capture
-
-### Search Options
-
-- **Result Count**: Configure number of search results
-- **Site Filtering**: Limit searches to specific domains
-- **Image Retention**: Include or exclude image content
-- **Alt Text Generation**: AI-generated image descriptions
-
-### Fact-Checking Modes
-
-- **Standard**: Quick factuality assessment
-- **Deep Dive**: Comprehensive analysis with multiple sources
-- **Evidence Scoring**: Quantified support/contradiction metrics
-
-## Use Cases
-
-### Research and Analysis
-Extract content from academic papers, news articles, and research websites for comprehensive analysis.
-
-### Content Verification
-Fact-check claims and statements using multiple web sources and credibility scoring.
-
-### Code Documentation
-Extract and analyze code from GitHub repositories and technical documentation sites.
-
-### Market Research
-Search for industry information, competitor analysis, and market trends across specific domains.
-
-### News Monitoring
-Track breaking news, extract article content, and verify information across multiple sources.
-
-### Academic Writing
-Gather sources, extract citations, and verify facts for research papers and articles.
-
-## Security Considerations
-
-**Third-Party Packages:**
-- These are community-maintained, not official Jina AI packages
-- Review source code on GitHub before installation
-- Monitor package updates and community feedback
-- Consider package reputation and maintenance activity
-
-**API Key Security:**
-- Store API keys securely in environment variables
-- Never commit API keys to version control
-- Monitor API usage for unusual activity
-- Rotate keys periodically
-
-**Content Filtering:**
-- Be aware that extracted content reflects source material
-- Implement additional filtering for sensitive applications
-- Consider content validation for critical use cases
-
-## Troubleshooting
-
-**Installation Issues:**
-- Verify Node.js installation and version
-- Check npm/npx permissions and configuration
-- Try clearing npm cache: `npm cache clean --force`
-
-**Server Connection Problems:**
-- Confirm MCP server shows as active
-- Check API key format and validity
-- Restart Jan after configuration changes
-- Verify network connectivity to Jina APIs
-
-**Content Extraction Failures:**
-- Some websites block automated access
-- Try different extraction modes (advanced server)
-- Check if target site requires authentication
-- Verify URL accessibility in browser
-
-**Rate Limiting:**
-- Jina AI has rate limits for free accounts
-- Consider upgrading API plan for higher limits
-- Implement delays between requests
-- Monitor usage through Jina dashboard
-
-**Tool Calling Issues:**
-- Ensure model supports tool calling
-- Check that Jina tools appear in tools panel
-- Try Claude 3.5+ Sonnet or GPT-4o for best results
-- Verify MCP permissions are enabled
-
-<Callout type="info">
-For technical issues with specific implementations, check the GitHub repositories for documentation and open issues.
-</Callout>
-
-## Server Comparison
-
-| Feature | Basic Server | Advanced Server |
-|---------|--------------|-----------------|
-| **Web Search** | ✓ | ✓ |
-| **Content Extraction** | ✓ | ✓ |
-| **Fact Checking** | ✗ | ✓ |
-| **Multiple Formats** | Limited | Full |
-| **Site Filtering** | ✓ | ✓ |
-| **Image Handling** | Basic | Advanced |
-| **GitHub Integration** | ✓ | ✓ |
-| **API Key Required** | Optional | Required |
-
-## Next Steps
-
-Jina MCP servers provide powerful web search and content extraction capabilities within Jan's privacy-focused environment. The community implementations offer different feature sets depending on your needs.
-
-For production use, consider contributing to these community projects or implementing additional security measures around third-party package usage.

docs/src/pages/docs/mcp-examples/search/perplexity.mdx

@@ -1,291 +0,0 @@
----
-title: Perplexity MCP
-description: Real-time web search and research capabilities through Perplexity's official MCP server.
-keywords:
-  [
-    Jan,
-    MCP,
-    Model Context Protocol,
-    Perplexity,
-    web search,
-    real-time search,
-    research,
-    Sonar API,
-    tool calling,
-  ]
----
-
-import { Callout, Steps } from 'nextra/components'
-
-# Perplexity MCP
-
-[Perplexity MCP](https://docs.perplexity.ai/guides/mcp-server) brings real-time web search directly into Jan through Perplexity's Sonar API. This official implementation lets AI models perform live web searches and return current, relevant information without the knowledge cutoff limitations of base models.
-
-Unlike static training data, this integration provides access to current web information, making it useful for research, fact-checking, and staying current with recent developments.
-
-## Available Tools
-
-### perplexity_ask
-- **Real-time web search**: Query current web information
-- **Conversational interface**: Multi-turn research conversations
-- **Source attribution**: Results include reference links
-- **Current information**: No knowledge cutoff restrictions
-
-## Prerequisites
-
-- Jan with MCP enabled
-- Perplexity API key from [perplexity.ai](https://perplexity.ai)
-- Model with tool calling support
-- Node.js installed (for NPX)
-
-<Callout type="info">
-This is Perplexity's official MCP implementation, providing direct access to their Sonar API infrastructure.
-</Callout>
-
-## Setup
-
-### Enable MCP
-
-1. Go to **Settings** > **MCP Servers**
-2. Toggle **Allow All MCP Tool Permission** ON
-
-![MCP settings page with toggle enabled](../../_assets/mcp-on.png)
-
-### Get Perplexity API Key
-
-1. Visit [perplexity.ai](https://perplexity.ai)
-2. Sign up for a Sonar API account
-3. Navigate to the developer dashboard
-4. Generate your API key
-5. Save the key
-
-[PLACEHOLDER: Screenshot of Perplexity dashboard showing API key generation]
-
-### Configure MCP Server
-
-Click `+` in MCP Servers section:
-
-**Configuration:**
-- **Server Name**: `perplexity-ask`
-- **Command**: `npx`
-- **Arguments**: `-y server-perplexity-ask`
-- **Environment Variables**: 
-  - Key: `PERPLEXITY_API_KEY`, Value: `your-api-key`
-
-[PLACEHOLDER: Screenshot of Jan MCP server configuration form with Perplexity settings]
-
-<Callout type="info">
-Using NPX means Jan handles package installation automatically. No need to clone repositories or manage dependencies manually.
-</Callout>
-
-### Verify Setup
-
-Check server status shows as active in the MCP Servers list.
-
-[PLACEHOLDER: Screenshot showing active Perplexity MCP server in Jan]
-
-### Model Configuration
-
-Use a tool-enabled model:
-
-- **Anthropic Claude 3.5+ Sonnet**
-- **OpenAI GPT-4o**
-- **Google Gemini Pro**
-
-[PLACEHOLDER: Screenshot showing model selection with tools enabled]
-
-## Usage
-
-Start a new chat with your tool-enabled model. Perplexity tools will appear in the available tools list.
-
-[PLACEHOLDER: Screenshot showing Perplexity tools in the tools panel]
-
-### Basic Web Search
-
-```
-What are the latest developments in quantum computing this week?
-```
-
-The AI will:
-1. Query Perplexity's search engine
-2. Return current, relevant information
-3. Include source links and references
-4. Provide up-to-date context
-
-### Research Queries
-
-```
-Compare the current market leaders in electric vehicle charging infrastructure and their recent partnerships.
-```
-
-This produces:
-- Current market analysis
-- Recent partnership announcements
-- Competitive landscape overview
-- Source attribution for claims
-
-### Real-Time Information
-
-```
-What's happening with the latest SpaceX launch? Any delays or updates?
-```
-
-Results include:
-- Current launch status
-- Recent updates or changes
-- Official announcements
-- News coverage links
-
-### Multi-Turn Research
-
-```
-Search for information about the new EU AI Act. Then tell me how it specifically affects small startups.
-```
-
-Enables:
-- Initial broad research
-- Follow-up targeted queries
-- Contextual understanding
-- Comprehensive analysis
-
-## Advanced Features
-
-### Conversational Research
-
-The `perplexity_ask` tool accepts conversation messages, enabling:
-- Multi-turn research sessions
-- Context-aware follow-up queries
-- Refined search based on previous results
-- Deeper exploration of topics
-
-### Source Attribution
-
-All results include:
-- Reference links to original sources
-- Publication dates when available
-- Authority indicators
-- Fact-checking context
-
-### Current Information Access
-
-Unlike base model training data:
-- No knowledge cutoff limitations
-- Real-time web information
-- Recent news and developments
-- Current market data
-
-## Use Cases
-
-### Journalism and Research
-Access current news, verify facts, and gather sources for articles and reports.
-
-### Market Intelligence
-Track industry developments, competitor announcements, and market trends in real-time.
-
-### Academic Research
-Find recent publications, current statistics, and up-to-date information for studies.
-
-### Business Analysis
-Monitor competitors, industry changes, and regulatory developments affecting your business.
-
-### Technical Research
-Stay current with framework updates, security patches, and technology announcements.
-
-### Investment Research
-Access current financial news, earnings reports, and market analysis for investment decisions.
-
-## Search Optimization
-
-### Effective Queries
-- **Be specific**: "Latest Tesla earnings Q4 2024" vs "Tesla news"
-- **Include timeframes**: "AI regulation changes this month"
-- **Specify sources**: "Recent academic papers on climate change"
-- **Ask follow-ups**: Build on previous searches for deeper insights
-
-### Research Strategies
-- Start broad, then narrow focus based on initial results
-- Use follow-up questions to explore specific aspects
-- Cross-reference multiple sources for verification
-- Request specific types of sources (academic, news, official)
-
-## Troubleshooting
-
-**API Key Issues:**
-- Verify API key format and validity
-- Check Perplexity account status and credits
-- Confirm API key permissions in dashboard
-- Test API key with direct API calls
-
-**Server Connection Problems:**
-- Ensure NPX can access npm registry
-- Check Node.js installation and version
-- Verify internet connectivity
-- Restart Jan after configuration changes
-
-**Search Quality Issues:**
-- Refine query specificity and context
-- Try different search approaches
-- Check if topic has recent information available
-- Verify sources are current and authoritative
-
-**Tool Calling Problems:**
-- Confirm model supports tool calling
-- Check that Perplexity tools appear in tools panel
-- Try Claude 3.5+ Sonnet for best results
-- Verify MCP permissions are enabled
-
-**Rate Limiting:**
-- Monitor API usage in Perplexity dashboard
-- Check account limits and quotas
-- Consider upgrading account plan if needed
-- Implement delays between searches if necessary
-
-<Callout type="warning">
-Perplexity API has usage limits based on your account plan. Monitor consumption to avoid service interruptions.
-</Callout>
-
-## Advanced Configuration
-
-### Custom Search Parameters
-
-The default implementation uses standard search parameters. For custom configurations:
-- Search model selection
-- Response length preferences
-- Source filtering options
-- Citation format preferences
-
-Modifications require editing the server configuration directly.
-
-### Docker Alternative
-
-For containerized environments, Perplexity also provides Docker configuration:
-```json
-{
-  "mcpServers": {
-    "perplexity-ask": {
-      "command": "docker",
-      "args": ["run", "-i", "--rm", "-e", "PERPLEXITY_API_KEY", "mcp/perplexity-ask"],
-      "env": {
-        "PERPLEXITY_API_KEY": "YOUR_API_KEY_HERE"
-      }
-    }
-  }
-}
-```
-
-## Perplexity vs Other Search MCPs
-
-| Feature | Perplexity | Exa | Jina (Community) |
-|---------|------------|-----|------------------|
-| **Official Support** | ✓ | ✓ | Community |
-| **Real-time Search** | ✓ | ✓ | ✓ |
-| **Source Attribution** | ✓ | ✓ | Limited |
-| **API Reliability** | High | High | Variable |
-| **Search Quality** | AI-optimized | Semantic | Standard |
-| **Rate Limits** | Plan-based | Plan-based | Varies |
-
-## Next Steps
-
-Perplexity MCP transforms Jan from a knowledge-cutoff-limited assistant into a current, web-connected research tool. The integration provides access to real-time information while maintaining Jan's privacy-focused local processing for conversation and reasoning.
-
-This combination delivers the best of both worlds: secure local AI processing with access to current web information when needed.