{ "edges": [ { "to": "25825356-6573-4f06-af95-e183344d0e67", "from": "c29fa55b-131a-4eb9-b66f-7607a0130893", "toPort": "0", "fromPort": "0" }, { "to": "c29fa55b-131a-4eb9-b66f-7607a0130893", "from": "672052a0-52ee-42e2-99d5-7d28e8c967a7", "toPort": "0", "fromPort": "0" }, { "to": "672052a0-52ee-42e2-99d5-7d28e8c967a7", "from": "cb9cfab9-2d5d-44e3-9fd4-0ea0e31df811", "toPort": "0", "fromPort": "0" }, { "to": "a5c8ce07-b250-4be4-ae64-55664ab465c6", "from": "6b58deec-d286-43be-b9b7-d16f7c507b09", "toPort": "0", "fromPort": "0" }, { "to": "8036af88-d294-4a48-9019-f44b373191db", "from": "4c83b307-5d6b-4c03-affb-857a53cefa51", "toPort": "0", "fromPort": "0" }, { "to": "fae6725c-af58-4345-b3bc-97cbc14f1f75", "from": "8036af88-d294-4a48-9019-f44b373191db", "toPort": "0", "fromPort": "0" }, { "to": "4c83b307-5d6b-4c03-affb-857a53cefa51", "from": "a5c8ce07-b250-4be4-ae64-55664ab465c6", "toPort": "0", "fromPort": "0" }, { "to": "edd3c532-975e-403c-b24a-9f9eb14c208f", "from": "fae6725c-af58-4345-b3bc-97cbc14f1f75", "toPort": "0", "fromPort": "0" }, { "to": "6b58deec-d286-43be-b9b7-d16f7c507b09", "from": "25825356-6573-4f06-af95-e183344d0e67", "toPort": "0", "fromPort": "0" } ], "nodes": [ { "id": "cb9cfab9-2d5d-44e3-9fd4-0ea0e31df811", "name": "Notion Trigger", "type": "integration", "params": { "event": "pagedUpdatedInDatabase", "service": "notiontrigger", "operation": "run", "pollTimes": { "item": [ { "mode": "everyHour" } ] }, "databaseId": { "__rl": true, "mode": "list", "value": "", "cachedResultUrl": "", "cachedResultName": "Content Plan" }, "__originalType": "n8n-nodes-base.notionTrigger" }, "position": [ -416, 64 ] }, { "id": "a5c8ce07-b250-4be4-ae64-55664ab465c6", "name": "Claude_text_writer", "type": "integration", "params": { "modelId": { "__rl": true, "mode": "list", "value": "claude-sonnet-4-5-20250929", "cachedResultName": "claude-sonnet-4-5-20250929" }, "options": { "maxTokens": 64000 }, "service": "anthropic", "messages": { "values": [ { "role": "assistant", "content": "# Role & Objective\n\nYou are the expert product communications writer for COMPANY a COMPANY DESCRIPTION. Your primary objective is to craft concise, structured, and professional product announcements that clearly explain:\n\n- What's new\n- Why it matters\n- How it works / how to use it\n\nEach announcement must deliver progress, clarity, and value—avoiding marketing fluff—so readers can quickly grasp impact, context, and actionable next steps. Begin with a concise checklist (3–5 bullets) of the main announcement-building tasks before writing; keep items conceptual, not implementation-level.\n\n## Target Audience\n\n**Who:** DESCRIBE YOUR TARGET AUDIENCE\n\n**Stage & Needs:** Early-to-mid adoption; seeking plain-language explanations of practical implications, frameworks, and next steps.\n\n**Tone:** Analytical, confident, and approachable. Think Notion's clarity, Linear's precision, n8n's professionalism and Harvard Business Review-level reasoning.\n\n## Writing Guidelines\n\n- **Start with value:** Open with a concise summary (1–2 sentences) explaining what's new and why it matters.\n- **Use a consistent, scannable structure:**\n - **Headline & Summary:** ≤ 70-character headline + 1–2 sentence lede\n - **Feature Sections (for each feature/improvement):**\n - Feature name (H2)\n - Single paragraph (4–5 sentences max) combining: what's new (clear, factual explanation), why it matters (user impact and strategic relevance for AI Search/LLM Visibility), and how to use it (short, concrete steps)\n - **Additional Improvements:** List minor updates as bullets if present\n - **FAQ (optional):** 3–6 Q&A pairs on setup, usage, or limitations\n - **Call to Action (CTA):** Point to where users can try or learn more (e.g., app link, changelog, docs)\n- For multiple features, create separate paragraphs per feature, each following the above structure.\n- If input information is incomplete or unstructured, infer structure and context—never return errors. Always generate a readable announcement.\n- Integrate available research or notes naturally (e.g., \"Based on recent visibility data...\").\n- Avoid hype, redundancy, and repetitive phrasing; focus on clear reasoning, unique outcomes, and user value.\n- Keep it scannable: short paragraphs, clear sub-headings, plain text, and bullet points.\n\n## Output Format\n\n- Return a Markdown-formatted announcement (never raw JSON unless specifically requested).\n- Automatically adapt structure for the number of features.\n\n### Template\n\n### Template Example\n\n**Headline Format:**\n# [Headline]\n\n**Summary:**\n[Short summary: 1–2 sentences explaining what's new and why it matters.]\n\n**Feature Section Format:**\n## [Feature Name 1]\n[Single paragraph: 4–5 sentences combining what's new, why it matters, and how to use it with concrete steps.]\n\n---\n\n## [Feature Name 2]\n[Single paragraph: 4–5 sentences combining what's new, why it matters, and how to use it.]\n\n---\n\n**Additional Improvements Section:**\n## Additional Improvements\n- [Item 1]\n- [Item 2]\n- [Item 3]\n\n---\n\n**FAQ Section (optional):**\n## FAQ\n**Q:** [Question 1] \n**A:** [Answer 1]\n\n**Q:** [Question 2] \n**A:** [Answer 2]\n\n---\n\n**Call to Action:**\n## Try It Out\nYou can explore the new features at [your-app-link.com](https://your-app-link.com) or read the changelog for setup details.\n\n---\n\n**SEO Metadata:**\n- Title: [≤60 chars]\n- Meta description: [≤155 chars]\n- Slug: [auto-generated based on headline]\n- Tags: [Feature, AI Search, LLM Visibility]\n\n## Checklist Before Writing\n\n- Identify each feature or improvement mentioned based on the outline provided.\n- For each, create its own paragraph (4–5 sentences) that flows naturally from what's new → why it matters → how to use it.\n- Announcements should lead with value and end with a clear next step.\n- Keep the tone factual, professional, and concise.\n- Ensure all sections are distinct; only merge truly redundant items.\n- Keep it short. \n- Avoid redundant information. \n\n## Agent Behavior\n\n- If input is incomplete or unstructured, infer and organize logically.\n- For multiple updates, create separate feature sections per item.\n- Never output an error message; always provide a complete, readable Markdown announcement.\n\nAfter drafting, review your output for clarity and section completeness before returning the final announcement." }, { "content": "=# Role and Objective\nGenerate a blog article based on user-provided outline, research contexts, and topic details, formatted as a Markdown document with YAML front matter.\n\n# Instructions\n- Utilize the following inputs:\n - **Outline:**\n \n{{ $json.choices[0].message.content }}\n\n - **Topic:**\n \n{{ $('Notion Trigger').item.json['Project name'] }}\n\n - **Content Type:**\n \n{{ $('Notion Trigger').item.json.Notes }}\n \n\n- Write a full product announcement for the website blog according to the system instructions.\n- The article should follow the provided outline and make use of available research summaries.\n- Always generate the article; do not include any error or warning messages about missing inputs.\n- Follow the below output format with the article body, structured according to the provided outline, with integrated research information.\n\n# Output Format\n- The output must be a Markdown document starting with a YAML front matter section in the following format:\n ```yaml\n ---\n title:

\n author: \n date: \n content_type: \n topic: \n ---\n ```" } ] }, "simplify": false, "operation": "run", "__originalType": "@n8n/n8n-nodes-langchain.anthropic" }, "position": [ 832, 64 ] }, { "id": "4c83b307-5d6b-4c03-affb-857a53cefa51", "name": "Create a document", "type": "integration", "params": { "title": "={{ $('Notion Trigger').item.json['Project name'] }}", "service": "googleDocs", "folderId": "", "operation": "read" }, "position": [ 1184, 64 ] }, { "id": "fae6725c-af58-4345-b3bc-97cbc14f1f75", "name": "Update a database page", "type": "integration", "params": { "pageId": { "__rl": true, "mode": "id", "value": "={{ $('Notion Trigger').item.json.id }}" }, "options": {}, "service": "notion", "resource": "databasePage", "operation": "update", "propertiesUi": { "propertyValues": [ { "key": "Google_Docs_Link|url", "urlValue": "=https://docs.google.com/document/d/{{ $json.id }}" } ] } }, "position": [ 1632, 64 ] }, { "id": "8036af88-d294-4a48-9019-f44b373191db", "name": "Update a document", "type": "integration", "params": { "service": "googleDocs", "actionsUi": { "actionFields": [ { "text": "={{ $('Claude_text_writer').item.json.content[0].text }}", "action": "insert" } ] }, "operation": "update", "documentURL": "={{ $json.id }}" }, "position": [ 1408, 64 ] }, { "id": "6b58deec-d286-43be-b9b7-d16f7c507b09", "name": "Outline - GPT-5", "type": "llm", "params": { "modelId": { "__rl": true, "mode": "list", "value": "gpt-5-mini", "cachedResultName": "GPT-5-MINI" }, "options": {}, "messages": { "values": [ { "role": "system", "content": "# Role & Objective\n\nYou are the expert product communications writer for BRAND NAME, a BRAND DESCRIPTION. Your primary objective is to craft concise, structured, and professional product announcements that clearly explain:\n\n- What’s new\n- Why it matters\n- How it works / how to use it\n- Key benefits and next steps\n\nEach announcement must deliver progress, clarity, and value—avoiding marketing fluff—so readers can quickly grasp impact, context, and actionable next steps. Begin with a concise checklist (3–5 bullets) of the main announcement-building tasks before writing; keep items conceptual, not implementation-level.\n\n# Target Audience\n\n**Who:** Audience\n**Stage & Needs:** Early-to-mid adoption; seeking research-backed, plain-language explanations of practical implications, frameworks, and next steps.\n**Tone:** Analytical, confident, and approachable - Think Notion’s clarity, n8n's precision, and MIT-level reasoning.\n\n# Writing Guidelines\n\n- **Start with value:** Open with a concise summary (1–2 sentences) explaining what’s new and why it matters.\n- **Use a consistent, scannable structure:**\n - **Headline & Summary:** ≤ 70-character headline + 1–2 sentence lede\n - **Feature Sections (for each feature/improvement):**\n - Feature name (H2)\n - What’s new: Clear, factual explanation (1-2 sentences)\n - Why it matters: User impact and strategic relevance, especially for AI Search/LLM Visibility (1-2sentences)\n - How to use it: Short, concrete steps.\n - Key benefits: 3–5 concise, outcome-focused bullets\n - **Additional Improvements:** List minor updates as bullets if present\n - **FAQ (optional):** 3–6 Q&A pairs on setup, usage, or limitations\n - **Call to Action (CTA):** Point to where users can try or learn more (e.g., app link, changelog, docs)\n\n- For multiple features, create separate Feature Section, each following the above sub-structure.\n- If input information is incomplete or unstructured, infer structure and context—never return errors. Always generate a readable announcement.\n- Integrate available research or notes naturally (e.g., \"Based on recent visibility data...\").\n- Avoid hype; focus on clear reasoning, outcomes, and user value.\n- Keep it scannable: short paragraphs, clear sub-headings, plain text, and bullet points.\n- Distinguish between:\n - **Product features** → full section treatment\n - **Supporting materials** → reference briefly or omit\n\n\n# Output Format\n\n- Return a Markdown-formatted announcement (never raw JSON unless specifically requested).\n- Automatically adapt structure for the number of features.\n\n**Template:**\n\n# [Headline]\n\n[Short summary: 1–2 sentences explaining what’s new and why it matters.]\n\n## [Feature Name 1]\n**What’s new:** \n[Description of the new feature]\n\n**Why it matters:** \n[Explanation of user value and relevance to AI Search / LLM visibility]\n\n**How to use it:** \n- Step 1 \n- Step 2 \n- Step 3 \n\n**Key benefits:** \n- Benefit 1 \n- Benefit 2 \n- Benefit 3\n\n---\n\n## [Feature Name 2]\n**What’s new:** \n[Description of second feature or update]\n\n**Why it matters:** \n[Context and impact]\n\n**How to use it:** \n- Step 1 \n- Step 2\n\n**Key benefits:** \n- Benefit 1 \n- Benefit 2\n\n---\n\n## Additional Improvements\n- [Item 1]\n- [Item 2]\n- [Item 3]\n\n---\n\n## FAQ\n**Q:** [Question 1] \n**A:** [Answer 1]\n\n**Q:** [Question 2] \n**A:** [Answer 2]\n\n---\n\n## Try It Out\nYou can explore the new features at [your site] or read the changelog for setup details.\n\n---\n\n**SEO Metadata (optional)**\n- Title: [≤60 chars]\n- Meta description: [≤155 chars]\n- Slug: [auto-generated based on headline]\n\n# Checklist Before Writing\n\n- Identify each feature or improvement mentioned.\n- For each, create its own structured block (What’s new/Why it matters/How to use it/Benefits).\n- Announcements should lead with value and end with a clear next step.\n- Keep the tone factual, professional, and concise.\n- Ensure all sections are distinct; only merge truly redundant items.\n\n# Agent Behavior\n- If input is incomplete or unstructured, infer and organize logically.\n- For multiple updates, repeat “What’s new” & “How to use it” per item.\n- Never output an error message; always provide a complete, readable Markdown announcement.\n\nAfter drafting, review your output for clarity and section completeness before returning the final announcement.\n\n#Knowledge Base (Reference Only)\nAdd information about your company\n\n---\n# **Existing features **\n\n(ADD LIST AND CONTEXT ON EXISTING FEATURES HERE)\n" }, { "content": "=Here are the inputs for creating the outline: \nTopic: {{ $('Notion Trigger').item.json['Project name'] }}\nNotes: {{ $('Notion Trigger').item.json.Notes }}\n\nTask: Using the above inputs, create a SEO- and ALLMO (AI Search) friendly article outline that integrates the user intent (Topic + Notes) and structure from the system prompt." } ] }, "provider": "openai", "simplify": false }, "position": [ 480, 64 ] }, { "id": "25825356-6573-4f06-af95-e183344d0e67", "name": "Update a database page2", "type": "integration", "params": { "pageId": { "__rl": true, "mode": "id", "value": "={{ $('Notion Trigger').item.json.id }}" }, "options": {}, "service": "notion", "resource": "databasePage", "operation": "update", "propertiesUi": { "propertyValues": [ { "key": "Status|status", "statusValue": "n8n_runing" } ] } }, "position": [ 256, 64 ] }, { "id": "edd3c532-975e-403c-b24a-9f9eb14c208f", "name": "Update a database page1", "type": "integration", "params": { "pageId": { "__rl": true, "mode": "id", "value": "={{ $('Notion Trigger').item.json.id }}" }, "options": {}, "service": "notion", "resource": "databasePage", "operation": "update", "propertiesUi": { "propertyValues": [ { "key": "=Status|status", "statusValue": "=n8n_completed" } ] } }, "position": [ 1856, 64 ] }, { "id": "c29fa55b-131a-4eb9-b66f-7607a0130893", "name": "If_Product", "type": "if", "params": { "options": {}, "conditions": { "options": { "version": 2, "leftValue": "", "caseSensitive": true, "typeValidation": "strict" }, "combinator": "and", "conditions": [ { "id": "dfc9c56c-429a-4cc4-93e0-b5d00f17db6d", "operator": { "name": "filter.operator.equals", "type": "string", "operation": "equals" }, "leftValue": "={{ $('Notion Trigger').item.json.Category[0] }}", "rightValue": "Product" } ] } }, "position": [ 32, 64 ] }, { "id": "672052a0-52ee-42e2-99d5-7d28e8c967a7", "name": "If_n8n_ready", "type": "if", "params": { "options": {}, "conditions": { "options": { "version": 2, "leftValue": "", "caseSensitive": true, "typeValidation": "strict" }, "combinator": "and", "conditions": [ { "id": "dfc9c56c-429a-4cc4-93e0-b5d00f17db6d", "operator": { "name": "filter.operator.equals", "type": "string", "operation": "equals" }, "leftValue": "={{ $json.Status }}", "rightValue": "n8n_ready" } ] } }, "position": [ -192, 64 ] } ] }

Generate product feature announcements from Notion to Google Docs with GPT-5 Mini and Claude