{ "edges": [ { "to": "4e20e15b-d178-4de7-b6dc-6ed91a8d0286", "from": "fe97c17b-ded9-4817-bae3-6a4da55ac3c2", "toPort": "0", "fromPort": "0" }, { "to": "5c2dc61b-77ca-44a8-9779-18643d20444e", "from": "6cc196e2-ad87-4442-b378-35cfa7fcd3d5", "toPort": "0", "fromPort": "0" }, { "to": "6cc196e2-ad87-4442-b378-35cfa7fcd3d5", "from": "5354e780-a6cb-4ca9-8e87-b8ef4aafc2a9", "toPort": "0", "fromPort": "0" }, { "to": "fe97c17b-ded9-4817-bae3-6a4da55ac3c2", "from": "5c2dc61b-77ca-44a8-9779-18643d20444e", "toPort": "0", "fromPort": "0" }, { "to": "fe97c17b-ded9-4817-bae3-6a4da55ac3c2", "from": "9bc878d7-29c1-4683-9499-d85e17955278", "toPort": "0", "fromPort": "ai_languageModel" }, { "to": "b2e47b17-01fb-496c-8a97-68bb51650d70", "from": "9bc878d7-29c1-4683-9499-d85e17955278", "toPort": "0", "fromPort": "ai_languageModel" }, { "to": "fe97c17b-ded9-4817-bae3-6a4da55ac3c2", "from": "b2e47b17-01fb-496c-8a97-68bb51650d70", "toPort": "0", "fromPort": "ai_outputParser" } ], "nodes": [ { "id": "5354e780-a6cb-4ca9-8e87-b8ef4aafc2a9", "name": "Error Trigger", "type": "trigger", "params": { "kind": "n8n-nodes-base.errorTrigger" }, "position": [ 0, 0 ] }, { "id": "6cc196e2-ad87-4442-b378-35cfa7fcd3d5", "name": "Set Context", "type": "set", "params": { "options": {}, "assignments": { "assignments": [ { "id": "cfg-001", "name": "n8n_instance_url", "type": "string", "value": "https://n8n.YOUR-DOMAIN.com" }, { "id": "cfg-002", "name": "n8n_api_key", "type": "string", "value": "YOUR_N8N_API_KEY_HERE" }, { "id": "cfg-003", "name": "telegram_chat_id", "type": "string", "value": "YOUR_TELEGRAM_CHAT_ID" }, { "id": "auto-001", "name": "error_workflow_id", "type": "string", "value": "={{ $('Error Trigger').item.json.workflow.id }}" }, { "id": "auto-002", "name": "error_workflow_name", "type": "string", "value": "={{ $('Error Trigger').item.json.workflow.name }}" }, { "id": "auto-003", "name": "error_node", "type": "string", "value": "={{ $('Error Trigger').item.json.execution.lastNodeExecuted }}" }, { "id": "auto-004", "name": "error_message", "type": "string", "value": "={{ $('Error Trigger').item.json.execution.error.message }}" }, { "id": "auto-005", "name": "error_stack", "type": "string", "value": "={{ $('Error Trigger').item.json.execution.error.stack }}" }, { "id": "auto-006", "name": "execution_id", "type": "string", "value": "={{ $('Error Trigger').item.json.execution.id }}" }, { "id": "auto-007", "name": "retry_of", "type": "string", "value": "={{ $('Error Trigger').item.json.execution.retryOf || '' }}" } ] } }, "position": [ 240, 0 ] }, { "id": "5c2dc61b-77ca-44a8-9779-18643d20444e", "name": "Get Workflow Content", "type": "http", "params": { "url": "={{ $('Set Context').item.json.n8n_instance_url + '/api/v1/workflows/' + $('Set Context').item.json.error_workflow_id }}", "options": {}, "sendHeaders": true, "headerParameters": { "parameters": [ { "name": "accept", "value": "application/json" }, { "name": "X-N8N-API-KEY", "value": "={{ $('Set Context').item.json.n8n_api_key }}" } ] } }, "position": [ 544, 0 ] }, { "id": "fe97c17b-ded9-4817-bae3-6a4da55ac3c2", "name": "AI Agent", "type": "llm", "params": { "text": "=Workflow Name: {{ $('Set Context').item.json.error_workflow_name }}\nFailed Node: {{ $('Set Context').item.json.error_node }}\nError Message: {{ $('Set Context').item.json.error_message }}\nStack Trace: {{ $('Set Context').item.json.error_stack }}\nRetry Of Execution: {{ $('Set Context').item.json.retry_of }}\nExecution Link: {{ $('Set Context').item.json.n8n_instance_url + '/workflow/' + $('Set Context').item.json.error_workflow_id + '/executions/' + $('Set Context').item.json.execution_id }}\nWorkflow Definition: {{ JSON.stringify($json) }}", "prompt": "=Workflow Name: {{ $('Set Context').item.json.error_workflow_name }}\nFailed Node: {{ $('Set Context').item.json.error_node }}\nError Message: {{ $('Set Context').item.json.error_message }}\nStack Trace: {{ $('Set Context').item.json.error_stack }}\nRetry Of Execution: {{ $('Set Context').item.json.retry_of }}\nExecution Link: {{ $('Set Context').item.json.n8n_instance_url + '/workflow/' + $('Set Context').item.json.error_workflow_id + '/executions/' + $('Set Context').item.json.execution_id }}\nWorkflow Definition: {{ JSON.stringify($json) }}", "options": { "systemMessage": "=You are an AI assistant specialized in analyzing n8n workflow errors. Your job is to generate a clear, actionable error report formatted for Telegram.\n\n## Error Classification\nClassify the error based on its message and stack trace:\n- HTTP 401/403 → \"🔐 Authentication Error\"\n- HTTP 404 → \"❌ Not Found Error\"\n- HTTP 429 → \"⏱️ Rate Limit Error\"\n- HTTP 500/502/503/504 → \"⚠️ Server Error\"\n- \"does not exist\" / \"credential\" → \"🔑 Credential Error\"\n- \"Cannot read\" / \"undefined\" / \"null\" → \"🐛 Undefined Property Error\"\n- \"Invalid JSON\" / \"parse\" → \"📄 JSON Parse Error\"\n- \"ECONNREFUSED\" / \"Connection\" / \"ETIMEDOUT\" → \"🔌 Connection Error\"\n- \"Timeout\" → \"⏰ Timeout Error\"\n- Other → \"🔴 Runtime Error\"\n\n## Output Format\nReturn ONLY a valid JSON object with a single field:\n{\n \"explanation\": \"Telegram HTML formatted string here\"\n}\n\n## Telegram HTML Rules\n- Allowed tags ONLY: , , , , \n- Use literal \\n for line breaks\n- No markdown, no other HTML tags\n- Do not use & < > characters outside of allowed HTML tags\n\n## Report Template (in English)\n🚨 N8N WORKFLOW ERROR\\n\\n📌 Workflow: {workflow_name}\\n⚙️ Failed Node: {failed_node}\\n🏷️ Error Type: {error_type}\\n\\n📋 Error Message:\\n{error_message_short}\\n\\n🔍 Root Cause:\\n{root_cause_1_2_sentences}\\n\\n✅ Fix Steps:\\n• {fix_step_1}\\n• {fix_step_2}\\n• {fix_step_3}\\n\\n{retry_line_only_if_retry_of_not_empty}\\n🔗 View execution details\n\n## Rules\n- MUST return ONLY the JSON object, no extra text before or after\n- Keep the explanation under 15 lines total\n- error_message_short: truncate to 200 characters if longer\n- Provide exactly 2-3 specific and actionable fix steps based on error type\n- If retry_of is empty or null, OMIT the retry line entirely\n- retry_line format: ⚠️ This is a retry of execution #{retry_of}\\n\\n\n- Use the workflow definition to give context-aware suggestions when possible" }, "provider": "openai", "promptType": "define", "approximated": true, "systemMessage": "=You are an AI assistant specialized in analyzing n8n workflow errors. Your job is to generate a clear, actionable error report formatted for Telegram.\n\n## Error Classification\nClassify the error based on its message and stack trace:\n- HTTP 401/403 → \"🔐 Authentication Error\"\n- HTTP 404 → \"❌ Not Found Error\"\n- HTTP 429 → \"⏱️ Rate Limit Error\"\n- HTTP 500/502/503/504 → \"⚠️ Server Error\"\n- \"does not exist\" / \"credential\" → \"🔑 Credential Error\"\n- \"Cannot read\" / \"undefined\" / \"null\" → \"🐛 Undefined Property Error\"\n- \"Invalid JSON\" / \"parse\" → \"📄 JSON Parse Error\"\n- \"ECONNREFUSED\" / \"Connection\" / \"ETIMEDOUT\" → \"🔌 Connection Error\"\n- \"Timeout\" → \"⏰ Timeout Error\"\n- Other → \"🔴 Runtime Error\"\n\n## Output Format\nReturn ONLY a valid JSON object with a single field:\n{\n \"explanation\": \"Telegram HTML formatted string here\"\n}\n\n## Telegram HTML Rules\n- Allowed tags ONLY: , , , , \n- Use literal \\n for line breaks\n- No markdown, no other HTML tags\n- Do not use & < > characters outside of allowed HTML tags\n\n## Report Template (in English)\n🚨 N8N WORKFLOW ERROR\\n\\n📌 Workflow: {workflow_name}\\n⚙️ Failed Node: {failed_node}\\n🏷️ Error Type: {error_type}\\n\\n📋 Error Message:\\n{error_message_short}\\n\\n🔍 Root Cause:\\n{root_cause_1_2_sentences}\\n\\n✅ Fix Steps:\\n• {fix_step_1}\\n• {fix_step_2}\\n• {fix_step_3}\\n\\n{retry_line_only_if_retry_of_not_empty}\\n🔗 View execution details\n\n## Rules\n- MUST return ONLY the JSON object, no extra text before or after\n- Keep the explanation under 15 lines total\n- error_message_short: truncate to 200 characters if longer\n- Provide exactly 2-3 specific and actionable fix steps based on error type\n- If retry_of is empty or null, OMIT the retry line entirely\n- retry_line format: ⚠️ This is a retry of execution #{retry_of}\\n\\n\n- Use the workflow definition to give context-aware suggestions when possible", "hasOutputParser": true }, "position": [ 784, 0 ] }, { "id": "9bc878d7-29c1-4683-9499-d85e17955278", "name": "Google Gemini Chat Model", "type": "llm", "params": { "options": {}, "provider": "google" }, "position": [ 704, 240 ] }, { "id": "b2e47b17-01fb-496c-8a97-68bb51650d70", "name": "Structured Output Parser", "type": "passthrough", "params": { "mode": "noOp", "autoFix": true, "jsonSchemaExample": "{\n \"explanation\": \"🚨 LỖI WORKFLOW N8N\\n\\n📌 Workflow: My Workflow\\n⚙️ Node lỗi: HTTP Request\"\n}" }, "position": [ 944, 240 ] }, { "id": "4e20e15b-d178-4de7-b6dc-6ed91a8d0286", "name": "Send Telegram Notification", "type": "integration", "params": { "text": "={{ $json.output.explanation }}", "chatId": "={{ $('Set Context').item.json.telegram_chat_id }}", "service": "telegram", "operation": "read", "additionalFields": { "parse_mode": "HTML", "appendAttribution": false } }, "position": [ 1168, 0 ] } ] }

Monitor workflow errors via n8n API with Gemini analysis and Telegram alerts