You are helping me build the following AI agent workflow.

## Goal
Triathlon Coach Analyzes Strava Data: After every Strava swim, bike, or run, the coach analyzes metrics like SWOLF and cadence to send instant training advice via email or...

## Specification
- What it does: Get personalized coaching for your swimming, cycling, and running sessions without manual data entry. The system automatically analyzes your Strava metrics like SWOLF, cadence, and heart rate zones to identify performance gaps. You receive actionable training recommendations and motivational advice directly via email or WhatsApp immediately after every activity.
- Trigger: Triggered by an event (Event · on new Strava activity)
- Autonomy: Fully hands-off
- Expected setup effort: about an afternoon
- Tools/services involved:
  - Strava Trigger: captures real-time activity updates
  - Code: preprocesses and flattens raw data
  - Google Gemini Chat Model: analyzes metrics and generates insights
  - Send Email: delivers detailed training advice
  - WhatsApp Business Cloud: sends instant summary notifications

## Known pitfalls, handle each one explicitly in your implementation
1. Latency spikes might delay the motivational message, so cache recent activities locally before querying the model.
2. HTML rendering varies across email clients, use a neutral template engine to ensure consistent display.
3. Personal data leakage is possible if prompts are too generic, include strict system instructions to never output raw user names.

## Reference implementation
https://n8n.io/workflows/2790 (workflow template)
Fetch and inspect this before building. If it matches my stack, adapt it;
if not, rebuild the pattern with my tools.

## Process requirements
1. Before building: ask me which of the listed tools I actually use and
   what my platform is (n8n / Make / code / other). Do not assume.
2. Adapt the pattern to my answers; do not force the reference stack.
3. Address every pitfall above; tell me how you handled each.
4. Provide a test plan I can run before letting this touch real data.
5. Ask before any step that sends messages, modifies data, or spends money.

Source: https://usecasesforagents.com/use-case/ai-triathlon-coach-analyzes-strava-data/ via usecasesforagents.com