> ## Agent Instructions > > Base URL: https://api.anysite.io > Authentication: send the `access-token` header. Do NOT use `Authorization: Bearer`. > Full endpoint catalog: https://app.anysite.io/docs # n8n MCP Tool Examples Real-world workflow examples using Anysite MCP in n8n ## Overview These examples demonstrate how to use the Anysite MCP Tool in n8n workflows to extract and process social media data. ## Example 1: LinkedIn Profile Enrichment Automatically enrich contact data with LinkedIn profile information. ### Workflow Structure ```mermaid graph LR A[Webhook Trigger] --> B[MCP Client] B --> C[AI Agent] C --> D[Process Data] D --> E[Store in Database] ``` ### Configuration #### Setup Webhook Trigger Configure a webhook that accepts LinkedIn profile URLs: ```json { "linkedin_url": "https://linkedin.com/in/username" } ``` #### Configure MCP Client - Add MCP Client node - Set endpoint to your Anysite Direct URL - Include `linkedin_user` tool #### Add AI Agent Node Configure the AI agent to use the MCP tool: ``` Extract detailed profile information from: {{ $json.linkedin_url }} Use the linkedin_user tool to get: - Full name and headline - Current position and company - Experience history - Education - Skills ``` #### Process Results Use a Code node to structure the extracted data: ```javascript const profile = $input.item.json; return { name: profile.full_name, headline: profile.headline, current_company: profile.current_position?.company, location: profile.location, skills: profile.skills, profile_url: profile.linkedin_url }; ``` ### Use Cases - CRM data enrichment - Lead qualification - Candidate screening - Contact database updates --- ## Example 2: Reddit Content Monitoring Monitor Reddit posts and extract detailed information for content analysis. ### Workflow Structure ```mermaid graph LR A[Schedule Trigger] --> B[Reddit URLs] B --> C[MCP Client] C --> D[AI Agent] D --> E[Sentiment Analysis] E --> F[Send Alert] ``` ### Configuration #### Setup Schedule Trigger Run every hour to check new Reddit posts: ``` Cron: 0 * * * * ``` #### Provide Reddit URLs Use a Set node with target post URLs or feed from a database. #### Configure MCP Client - Include `reddit_post` tool - Set for processing multiple items #### Extract Post Data AI agent prompt: ``` Analyze this Reddit post: {{ $json.reddit_url }} Extract: - Post title and content - Author information - Upvotes and comments count - Top comments - Post timestamp ``` #### Analyze Sentiment Add another AI node to analyze sentiment: ``` Analyze the sentiment of this post and its top comments. Classify as: Positive, Negative, or Neutral Extract key topics and themes. ``` ### Use Cases - Brand mention monitoring - Community sentiment tracking - Competitor analysis - Trend identification --- ## Example 3: Instagram Profile Analysis Extract and compare Instagram profiles for influencer research. ### Workflow Configuration #### Input Instagram URLs Webhook or manual trigger with profile URLs: ```json { "profiles": [ "https://instagram.com/influencer1", "https://instagram.com/influencer2" ] } ``` #### MCP Client Setup - Include `instagram_profile` tool - Enable batch processing #### Extract Profile Data AI agent instruction: ``` For each Instagram profile, extract: - Username and full name - Follower count and engagement rate - Bio and contact information - Recent posts count - Account type (business/personal) ``` #### Compare and Rank Use Code node to compare profiles: ```javascript const profiles = $input.all().map(item => item.json); // Calculate engagement scores profiles.forEach(profile => { profile.engagement_score = (profile.avg_likes + profile.avg_comments) / profile.followers * 100; }); // Sort by engagement profiles.sort((a, b) => b.engagement_score - a.engagement_score ); return profiles; ``` ### Use Cases - Influencer selection - Campaign planning - Competitive analysis - Audience research --- ## Example 4: Multi-Platform Data Aggregation Collect data from multiple social platforms for comprehensive analysis. ### Workflow Overview Combine LinkedIn, Instagram, and Reddit data for a person or brand. #### Setup Input Provide multiple social media URLs: ```json { "linkedin": "https://linkedin.com/in/username", "instagram": "https://instagram.com/username", "reddit": "https://reddit.com/user/username" } ``` #### Parallel Processing Split workflow into parallel branches: - Branch 1: LinkedIn data extraction - Branch 2: Instagram data extraction - Branch 3: Reddit data extraction Each branch uses MCP Client with appropriate tool. #### Merge Results Use Merge node to combine all data: ``` Mode: Combine All Output: Single item with all social data ``` #### Generate Report AI agent creates unified analysis: ``` Based on the social media data from LinkedIn, Instagram, and Reddit: 1. Summarize professional background 2. Analyze content themes and interests 3. Evaluate audience engagement 4. Identify key insights and patterns ``` ### Use Cases - Comprehensive background checks - Brand presence analysis - Content strategy research - Cross-platform insights --- ## Example 5: Automated LinkedIn Company Research Research companies automatically from a list. ### Configuration #### Input Company List Provide company LinkedIn URLs via webhook or spreadsheet: ```json { "companies": [ "https://linkedin.com/company/company1", "https://linkedin.com/company/company2" ] } ``` #### MCP Client Use `linkedin_company` tool for data extraction. #### Extract Company Data ``` For each company, extract: - Company name and industry - Size and location - Description and specialties - Employee count - Recent updates and posts ``` #### Export Results Format and export to: - Google Sheets - Airtable - CSV file - Database ### Use Cases - Market research - Lead generation - Competitive intelligence - Partnership opportunities --- ## Best Practices #### Error Handling Add Error Trigger nodes to handle API failures gracefully and implement retry logic. #### Rate Limiting Use Wait nodes between MCP calls to avoid hitting API rate limits. #### Data Validation Validate extracted data before processing to ensure quality and completeness. #### Logging Log all MCP operations for debugging and audit purposes. ## Tips for Optimization 1. **Batch Processing**: Group multiple URLs together to reduce workflow execution time 2. **Caching**: Store frequently accessed data to minimize API calls 3. **Parallel Execution**: Use Split In Batches node for processing large datasets 4. **Error Recovery**: Implement fallback mechanisms for failed API calls 5. **Monitoring**: Set up notifications for workflow failures or data anomalies ## Advanced Patterns ### Pattern 1: Conditional Tool Selection Use IF nodes to dynamically select which MCP tool to use based on URL type: ```javascript // Detect platform from URL const url = $json.social_url; if (url.includes('linkedin.com/in/')) { return { tool: 'linkedin_user' }; } else if (url.includes('linkedin.com/company/')) { return { tool: 'linkedin_company' }; } else if (url.includes('instagram.com')) { return { tool: 'instagram_profile' }; } ``` ### Pattern 2: Incremental Updates Track and update only changed data: ```javascript // Compare with existing data const existing = $('Database').first().json; const fresh = $json; const changes = {}; if (existing.followers !== fresh.followers) { changes.followers = fresh.followers; changes.follower_growth = fresh.followers - existing.followers; } return changes.followers ? changes : null; ``` ### Pattern 3: Data Enrichment Pipeline Chain multiple MCP tools for comprehensive enrichment: ``` LinkedIn User → Get Company → Get Company Employees → Analyze Network ``` ## Resources - [View All Available MCP Tools](/docs/mcp-server/tools) - [n8n MCP Tool Installation](/docs/mcp-server/n8n-tool/installation) - [n8n Official Documentation](https://docs.n8n.io/) - [Anysite API Reference](/docs/api) ## Need Help? #### Get Support Contact our support team for assistance with n8n MCP integration