{ "name": "HubSpot to LinkedIn Auto-Connect With GPT4o Message", "flow": [ { "id": 2, "module": "hubspotcrm:getContact", "version": 2, "parameters": { "__IMTCONN__": 1912907 }, "mapper": { "contactId": "{ID}", "parseCustomFields": true }, "metadata": { "designer": { "x": -270, "y": 11, "name": "Fetch HubSpot Contact Data" }, "restore": { "parameters": { "__IMTCONN__": { "label": "My HubSpot CRM connection (contact@justvito.io)", "data": { "scoped": "true", "connection": "hubspotcrm" } } }, "expect": { "contactId": { "mode": "edit" }, "outputProperties": { "mode": "chose" }, "parseCustomFields": { "mode": "chose" } } }, "parameters": [ { "name": "__IMTCONN__", "type": "account:hubspotcrm", "label": "Connection", "required": true } ], "expect": [ { "name": "contactId", "type": "select", "label": "Contact ID", "required": true }, { "name": "outputProperties", "type": "select", "label": "Output Properties", "multiple": true }, { "name": "parseCustomFields", "type": "boolean", "label": "Parse Custom Fields", "required": true } ] } }, { "id": 5, "module": "openai-gpt-3:CreateCompletion", "version": 1, "parameters": { "__IMTCONN__": 1796128 }, "mapper": { "select": "chat", "temperature": ".5", "top_p": "1", "n_completions": "1", "frequency_penalty": ".3", "presence_penalty": ".4", "model": "chatgpt-4o-latest", "max_tokens": "100", "messages": [ { "role": "assistant", "content": "You are a messaging assistant that writes short, personalized LinkedIn connection request notes. Your output must be concise, relevant, and 100% focused on the recipient - not on pitching, selling, or talking about yourself.\n\nThe tone should be friendly and direct, with no fluff or generic phrases. Highlight a shared interest, role, or reason for connecting based on the person’s job title, industry, or company.\n\nOutput only the message - do not include greetings like \"Hi\" or \"Hello\" and do not add sign-offs like \"Thanks\" or \"Best.\"\n\n🧾 Input Example (Prompt Template):\n\nWrite a short and personal LinkedIn connection note based on this person's role and company:\n\n- Name: Sarah Johnson\n- Title: Director of Product Marketing\n- Company: Typeform\n- Reason: We're both focused on simplifying customer-facing experiences with automation and personalized UX.\n\n✅ Example Output:\nReally admire how Typeform approaches user-friendly workflows -your work in product marketing must play a huge role in that. Would love to connect.\n\n🔐 DATA REFERENCE FOR PERSONALIZATION\n- First Name: '{{2.properties.firstname}},'\n- Positioning Snapshot (Company): '{positioning_snapshot},'\n- Site Summary & Buyer Insights (Company): 'site_summary_and_buyer_insights'\n- External Signal Scanner (Company): '{external_signal_scanner}'\n- Company Name: '{{2.properties.company}}'\n- Title: '{{2.properties.jobtitle}},'\n- Professional Summary: '{professional_summary},'\n- Headline: '{headline},'\n- Key Skills: 'key_skills},'\n- Role Focus: '{role_focus},'" } ], "response_format": "text" }, "metadata": { "designer": { "x": 121, "y": 9, "name": "Generate Short LinkedIn Connect Message (GPT-4o)" }, "restore": { "parameters": { "__IMTCONN__": { "label": "My OpenAI connection", "data": { "scoped": "true", "connection": "openai-gpt-3" } } }, "expect": { "select": { "label": "Create a Chat Completion (GPT and o1 models)" }, "logit_bias": { "mode": "chose" }, "tool_choice": { "mode": "chose", "label": "Empty" }, "stop": { "mode": "chose" }, "additionalParameters": { "mode": "chose" }, "model": { "mode": "chose", "label": "chatgpt-4o-latest (system)" }, "messages": { "mode": "chose", "items": [ { "role": { "mode": "chose", "label": "Assistant" }, "tool_calls": { "mode": "edit" } } ] }, "response_format": { "mode": "chose", "label": "Text" } } }, "parameters": [ { "name": "__IMTCONN__", "type": "account:openai-gpt-3", "label": "Connection", "required": true } ], "expect": [ { "name": "select", "type": "select", "label": "Select Method", "required": true, "validate": { "enum": [ "chat", "prompt" ] } }, { "name": "temperature", "type": "number", "label": "Temperature", "validate": { "max": 2, "min": 0 } }, { "name": "top_p", "type": "number", "label": "Top P", "validate": { "max": 1, "min": 0 } }, { "name": "n_completions", "type": "number", "label": "Number" }, { "name": "frequency_penalty", "type": "number", "label": "Frequency Penalty", "validate": { "max": 2, "min": -2 } }, { "name": "presence_penalty", "type": "number", "label": "Presence Penalty", "validate": { "max": 2, "min": -2 } }, { "name": "logit_bias", "type": "array", "label": "Token Probability", "spec": { "spec": [ { "name": "token", "type": "text", "label": "Token ID", "required": true }, { "name": "probability", "type": "number", "label": "Probability", "required": true, "validate": { "max": 100, "min": -100 } } ], "type": "collection", "label": "Token Probability", "name": "value" } }, { "name": "seed", "type": "integer", "label": "Seed" }, { "name": "tool_choice", "type": "select", "label": "Tool Choice", "validate": { "enum": [ "none", "auto", "required" ] } }, { "name": "stop", "type": "array", "label": "Stop Sequences", "validate": { "maxItems": 4 }, "spec": { "type": "text", "label": "Stop Sequence", "name": "value" } }, { "name": "additionalParameters", "type": "array", "label": "Other Input Parameters", "spec": { "spec": [ { "name": "key", "type": "text", "label": "Parameter Name", "required": true }, { "name": "type", "type": "select", "label": "Input Type", "options": [ { "label": "Text", "value": "text", "nested": [ { "name": "value", "type": "text", "label": "Parameter Value" } ], "default": true }, { "label": "Number", "value": "number", "nested": [ { "name": "value", "type": "number", "label": "Parameter Value" } ] }, { "label": "Boolean", "value": "boolean", "nested": [ { "name": "value", "type": "boolean", "label": "Parameter Value" } ] }, { "label": "Date", "value": "date", "nested": [ { "name": "value", "type": "date", "label": "Parameter Value" } ] }, { "label": "Any", "value": "any", "nested": [ { "name": "value", "type": "any", "label": "Parameter Value" } ] } ] } ], "type": "collection", "label": "Input Parameter", "name": "value" } }, { "name": "model", "type": "select", "label": "Model", "required": true }, { "name": "max_tokens", "type": "uinteger", "label": "Max Completion Tokens" }, { "name": "messages", "type": "array", "label": "Messages", "required": true, "spec": { "label": "Message", "type": "collection", "spec": [ { "name": "role", "type": "select", "label": "Role", "required": true, "options": { "store": [ { "label": "User", "value": "user", "nested": [ { "help": "Text content of the message on behalf of the selected __Role__.", "name": "content", "type": "text", "label": "Text Content" }, { "name": "imageInputType", "type": "select", "label": "Image Input Type", "mappable": false, "options": [ { "value": "url", "label": "URL", "nested": [ { "name": "imageUrl", "label": "Image URL", "type": "url", "help": "Make sure to use a publicly accessible URL.\nYou can test if your image is publicly accessible by opening the link in an incognito tab." } ] }, { "value": "file", "label": "Image File", "nested": [ { "name": "imageFile", "label": "Image", "type": "collection", "spec": [ { "name": "imageFilename", "type": "filename", "label": "Image Filename", "semantic": "file:name", "extension": [ "jpg", "jpeg", "png", "webp", "gif" ], "help": "Accepted extensions: `.jpg`, `.jpeg`, `.png`, `.webp` and `.gif`." }, { "name": "imageData", "type": "buffer", "label": "Image Data", "semantic": "file:data" } ] } ] } ] }, { "name": "imageDetail", "type": "select", "label": "Image Detail", "help": "Recommended value: `Auto`", "options": [ { "value": "auto", "default": true, "label": "Auto" }, { "value": "high", "label": "High" }, { "value": "low", "label": "Low" } ] } ] }, { "label": "Assistant", "value": "assistant", "nested": [ { "help": "Text content of the message on behalf of the selected __Role__.", "name": "content", "type": "text", "label": "Text Content" }, { "name": "tool_calls", "label": "Tool Calls", "type": "array", "labels": { "add": "Add tool call" }, "mode": "edit", "mappable": { "help": "You can map the entire `Choices[]: Message.Tool Calls` array from a previous Create a Completion module here." }, "spec": { "label": "Tool Call", "type": "collection", "spec": [ { "name": "type", "type": "hidden", "default": "function" }, { "name": "id", "type": "text", "label": "Tool call ID", "help": "Map this directly from the output of a previous **Create a Completion** module. Look for `Choices[]: Message.Tool Calls[]: ID`." }, { "name": "function", "type": "collection", "label": "Function", "spec": [ { "name": "name", "label": "Name", "type": "text", "required": true, "help": "The name of the function previously called." }, { "name": "arguments", "label": "Arguments", "type": "text", "required": true, "help": "The arguments previously output by the AI." } ] } ] } } ] }, { "label": "Developer / System", "nested": [ { "help": "Text content of the message on behalf of the selected __Role__.", "name": "content", "type": "text", "label": "Text Content" } ], "value": "system" }, { "label": "Tool", "value": "tool", "nested": [ { "help": "The return of the function. This role should only be used when you have processed a previous function call and want to send the output of the function execution back to the AI.", "name": "content", "type": "text", "label": "Text Content", "required": true }, { "label": "Tool Call ID.", "name": "tool_call_id", "type": "text", "required": true, "help": "Map this directly from the output of a previous **Create a Completion** module. Look for `Choices[]: Message.Tool Calls[]: ID`." } ] } ] } } ], "name": "value" } }, { "name": "response_format", "type": "select", "label": "Response Format", "validate": { "enum": [ "text", "json_object" ] } }, { "name": "prediction", "type": "text", "label": "Predicted Outputs" } ], "interface": [ { "name": "result", "type": "any", "label": "Result" }, { "name": "id", "type": "text", "label": "ID" }, { "name": "object", "type": "text", "label": "Object" }, { "name": "created", "type": "date", "label": "Created" }, { "name": "model", "type": "text", "label": "Model" }, { "name": "choices", "spec": [ { "name": "text", "type": "text", "label": "Text" }, { "name": "index", "type": "number", "label": "Index" }, { "name": "logprobs", "type": "text", "label": "Log Probs" }, { "name": "finish_reason", "type": "text", "label": "Finish Reason" }, { "name": "message", "spec": [ { "name": "role", "type": "text", "label": "Role" }, { "name": "content", "type": "text", "label": "Content" }, { "name": "tool_calls", "spec": [ { "name": "id", "type": "text", "label": "ID" }, { "name": "type", "type": "text", "label": "Type" }, { "name": "function", "spec": [ { "name": "name", "type": "text", "label": "Name" }, { "name": "arguments", "type": "text", "label": "Arguments" } ], "type": "collection", "label": "Function" } ], "type": "array", "label": "Tool Calls" }, { "name": "refusal", "type": "text", "label": "Refusal" }, { "name": "annotations", "spec": [ { "name": "type", "type": "text", "label": "Type" }, { "name": "url_citation", "spec": [ { "name": "end_index", "type": "number", "label": "End Index" }, { "name": "start_index", "type": "number", "label": "Start Index" }, { "name": "title", "type": "text", "label": "Title" }, { "name": "url", "type": "text", "label": "URL" } ], "type": "collection", "label": "URL Citation" } ], "type": "array", "label": "Annotations" } ], "type": "collection", "label": "Message" } ], "type": "array", "label": "Choices" }, { "name": "usage", "spec": [ { "name": "prompt_tokens", "type": "number", "label": "Prompt Tokens" }, { "name": "completion_tokens", "type": "text", "label": "Completion Tokens" }, { "name": "total_tokens", "type": "number", "label": "Total Tokens" }, { "name": "prompt_tokens_details", "spec": [ { "name": "cached_tokens", "type": "uinteger", "label": "Cached Tokens" }, { "name": "text_tokens", "type": "uinteger", "label": "Text Tokens" }, { "name": "image_tokens", "type": "uinteger", "label": "Image Tokens" }, { "name": "audio_tokens", "type": "uinteger", "label": "Audio Tokens" } ], "type": "collection", "label": "Prompt Tokens Details" }, { "name": "completion_tokens_details", "spec": [ { "name": "reasoning_tokens", "type": "uinteger", "label": "Reasoning Tokens" }, { "name": "text_tokens", "type": "uinteger", "label": "Text Tokens" }, { "name": "audio_tokens", "type": "uinteger", "label": "Audio Tokens" }, { "name": "accepted_prediction_tokens", "type": "uinteger", "label": "Accepted Prediction Tokens" }, { "name": "rejected_prediction_tokens", "type": "uinteger", "label": "Rejected Prediction Tokens" } ], "type": "collection", "label": "Completion Tokens Details" } ], "type": "collection", "label": "Usage" }, { "name": "service_tier", "type": "text", "label": "Service Tier" }, { "name": "system_fingerprint", "type": "text", "label": "System Fingerprint" } ], "advanced": true } }, { "id": 4, "module": "custom-js:runPuppeteer", "version": 1, "parameters": { "__IMTCONN__": 4076083, "returnValueType": "binary" }, "mapper": { "code": "module.exports = async ({ page, data }) => {\n const url = data.linkedinUrl; // Must be a full https:// LinkedIn profile URL\n const message = data.connectMessage; // Personalized message string from OpenAI\n\n // Navigate to the LinkedIn profile\n await page.goto(url, { waitUntil: 'networkidle2' });\n\n // Optional: Screenshot before interaction\n await page.screenshot({ path: '/tmp/debug_screenshot.png' });\n\n let connectClicked = false;\n\n // Try clicking the \"Connect\" or \"Invite\" button directly\n try {\n await page.waitForSelector('button[aria-label*=\"Connect\"], button[aria-label=\"Invite\"]', {\n visible: true,\n timeout: 5000\n });\n await page.click('button[aria-label*=\"Connect\"], button[aria-label=\"Invite\"]');\n connectClicked = true;\n } catch (error) {\n console.warn(`⚠️ 'Connect' button not found directly. Trying 'More actions' fallback...`);\n }\n\n // Fallback: click \"More actions\" → \"Connect\"\n if (!connectClicked) {\n try {\n await page.waitForSelector('button[aria-label=\"More actions\"]', { visible: true, timeout: 5000 });\n await page.click('button[aria-label=\"More actions\"]');\n\n await page.waitForSelector('div[role=\"menuitem\"]:has-text(\"Connect\")', { visible: true, timeout: 5000 });\n await page.click('div[role=\"menuitem\"]:has-text(\"Connect\")');\n } catch (error) {\n throw new Error(`❌ Could not find 'Connect' option on the profile: ${url}`);\n }\n }\n\n // Wait for the Add a note button and click it\n try {\n await page.waitForSelector('button[aria-label=\"Add a note\"]', { visible: true, timeout: 5000 });\n await page.click('button[aria-label=\"Add a note\"]');\n } catch (error) {\n throw new Error(`❌ 'Add a note' button not found. Cannot attach message.`);\n }\n\n // Wait for message text area and type the custom message\n try {\n await page.waitForSelector('textarea[name=\"message\"]', { visible: true, timeout: 5000 });\n await page.type('textarea[name=\"message\"]', message, { delay: 10 });\n } catch (error) {\n throw new Error(`❌ Failed to type message in textarea.`);\n }\n\n // Send the invitation with message\n try {\n await page.waitForSelector('button[aria-label=\"Send now\"], button[aria-label=\"Send invitation\"]', {\n visible: true,\n timeout: 5000\n });\n await page.click('button[aria-label=\"Send now\"], button[aria-label=\"Send invitation\"]');\n } catch (error) {\n throw new Error(`❌ 'Send invitation' button not found on the profile: ${url}`);\n }\n\n return { message: `✅ Connection request sent successfully with message to ${url}` };\n};\n​" }, "metadata": { "designer": { "x": 605, "y": 10, "name": "Send LinkedIn Connect w/ GPT Message Custom JS" }, "restore": { "parameters": { "__IMTCONN__": { "label": "My Custom JS V2 connection", "data": { "scoped": "true", "connection": "custom-js2" } }, "returnValueType": { "label": "Screenshot (PNG)" } } }, "parameters": [ { "name": "__IMTCONN__", "type": "account:custom-js2", "label": "Connection", "required": true }, { "name": "returnValueType", "type": "select", "label": "Return Type", "required": true, "validate": { "enum": [ "text", "binary" ] } } ], "expect": [ { "name": "code", "type": "text", "label": "Puppeteer Commands." } ], "interface": [ { "name": "output", "type": "binary", "label": "Output" } ] } } ], "metadata": { "instant": false, "version": 1, "scenario": { "roundtrips": 1, "maxErrors": 3, "autoCommit": true, "autoCommitTriggerLast": true, "sequential": false, "slots": null, "confidential": false, "dataloss": false, "dlq": false, "freshVariables": false }, "designer": { "orphans": [] }, "zone": "us1.make.com", "notes": [ { "moduleIds": [ 2 ], "content": "

Use a different CRM? Replace this module with your CRM's node.

Need help? Ping Nate Shelly

", "isFilterNote": false, "metadata": { "color": "#9138FE" } }, { "moduleIds": [ 4 ], "content": "

What does this JavaScript do?

Automates a LinkedIn connection request using a headless browser with a personalized message.

Requires a browser automation runtime (e.g., headless Chrome via Puppeteer, Browserless, or similar)

Notes

Want to modify the action?

Replace the JavaScript code inside this module to handle different interactions, scraping, or browser logic.

Need help? Ping Nate Shelly

", "isFilterNote": false, "metadata": { "color": "#9138FE" } }, { "moduleIds": [ 5 ], "content": "

Purpose:

Generates a concise and personalized LinkedIn connection note using GPT-4o. The message is focused entirely on the recipient - their role, company, and focus - with no sales pitch or mention of you.

Tone:

Friendly, direct, and relevant. No fluff, greetings, or sign-offs.

Prompt Logic:

The system prompt includes:

Inputs Used:

Output:

A clean, short message ready to inject into the browser-based JavaScript module that sends a LinkedIn connection request.

Need a different output or behavior?

You can update the system prompt, swap out any dynamic data fields listed under 🔐 DATA REFERENCE FOR PERSONALIZATION, or adjust settings like Max Tokens, Temperature, and other model parameters directly within this module. You can also add HTTP modules to connect with third-party APIs for scraping or enriching data from platforms like LinkedIn, X (Twitter), or other social sources using RapidAPI or Apify.

Need help? Ping Nate Shelly

", "isFilterNote": false, "metadata": { "color": "#9138FE" } } ] } }