Log workouts with n8n form and get Gemini AI fitness analysis

Fill a web form → Gemini AI parses workout → Save to Google Sheets → Slack + Gmail notifications

Who

Gym-goers, runners, home workout enthusiasts, and personal trainers who want to avoid complicated fitness apps

You get

Log workouts via a simple form; AI auto-parses exercises, muscle groups, calories, and intensity into a structured training log

You need

n8n · gemini · google-sheets · slack · gmail · code

Steps

1. Prerequisites and credential setup

   Create a Google Sheet with a "Workouts" tab and add column headers: Date, Workout, Duration, Feeling, Type, Exercises, Calories, Muscles, Intensity, Tips, Next Suggestion, Notes, Logged At. Add your Google Sheets and Gmail credentials in n8n. Get a free Gemini API key from Google AI Studio and configure it in n8n. Connect your Slack app and obtain the channel ID. Update the recipient email address in the Gmail node.

2. Fill the workout form and trigger the workflow

   Visit and submit the form URL generated by the n8n Form Trigger. Form fields: "What did you do today?" (required, textarea, e.g. "Bench press 60kg x10 x3, 5km run 28min"), "How do you feel? (1-10)" (required, dropdown), "Duration (minutes)" (required, number), "Notes" (optional, textarea for injuries or PRs). Submission automatically triggers the workflow.

3. Gemini AI analyzes the workout

   The workflow passes form data to the "Analyze Workout with AI" node (Basic LLM Chain), which calls the Google Gemini Chat Model (gemini-1.5-flash-latest, temperature=0.2). The node prompt instructs the AI to act as a certified personal trainer and sports scientist, analyze the workout log, and return ONLY valid JSON with no markdown or extra text.

   Execution prompt · Source text。n8n "Analyze Workout with AI" node (Basic LLM Chain) text parameter field

   Copy prompt

   You are a certified personal trainer and sports scientist. Analyze this workout log and return ONLY valid JSON (no markdown, no extra text):
   
   Workout description: {{ $json['What did you do today?'] }}
   Duration: {{ $json['Duration (minutes)'] }} minutes
   Feeling: {{ $json['How do you feel? (1-10)'] }}
   Notes: {{ $json['Notes'] || 'None' }}
   
   Return this exact JSON structure:
   {
     "exercises": [
       {
         "name": "Exercise name",
         "type": "strength/cardio/flexibility/other",
         "muscle_groups": ["primary muscles worked"],
         "sets": 0,
         "reps": 0,
         "weight_kg": 0,
         "duration_min": 0,
         "estimated_calories": 0
       }
     ],
     "summary": {
       "total_exercises": 0,
       "total_calories": 0,
       "primary_type": "strength/cardio/mixed",
       "muscle_groups_hit": ["list of all muscles worked"],
       "intensity": "low/moderate/high/very high"
     },
     "tips": [
       "One specific, actionable tip based on this workout",
       "Another tip about recovery or next session"
     ],
     "next_workout_suggestion": "Brief suggestion for complementary workout"
   }

4. Parse, validate, and save to Google Sheets

   The "Parse AI Results" Code node receives the AI response, extracts JSON (supports markdown code block parsing), and falls back to a default structure on failure. It also reads original form data and combines them into a structured record. Then the "Save to Workout Sheet" node appends the data to the Google Sheet with field mappings for Date, Workout, Duration, Feeling, Type, Exercises, Calories, Muscles, Intensity, Tips, Next Suggestion, Notes, Logged At. Replace YOUR_GOOGLE_SHEET_ID in the documentId parameter.

5. Build summary and send notifications

   The "Build Summary Message" Code node generates emoji-rich Slack messages and HTML email content based on workout type, intensity, and feeling. These are sent in parallel to a Slack channel (replace YOUR_SLACK_CHANNEL_ID) and via Gmail (replace YOUR_EMAIL_ADDRESS). Email subject format: "Workout logged: {type} {duration}min — {calories} kcal ({date})".

You are setting up an n8n workflow to collect workout logs via a form and auto-analyze them with Gemini AI. Please confirm: 1) Google Sheet created with "Workouts" tab and headers; 2) Google Sheets, Gmail, Slack credentials and Gemini API key configured in n8n; 3) workflow.json imported or nodes rebuilt; 4) all YOUR_* placeholders replaced. Then activate the workflow, test with a sample workout entry, and verify the Google Sheet row, Slack message, and Gmail summary.