You are helping me build the following AI agent workflow.

## Goal
Email-Based Scheduling Automation: Listens for email replies to schedule proposals, parses the intent via AI, and auto-creates calendar events or requests alternatives.

## Specification
- What it does: You save time on back-and-forth scheduling by letting this agent read your Gmail replies and instantly book confirmed slots in your Google Calendar or propose new times if needed. It handles the parsing of your free-text responses so you never have to manually create events again.
- Trigger: Triggered by an event (Event · on new email reply)
- Autonomy: Fully hands-off
- Expected setup effort: about an afternoon
- Tools/services involved:
  - Gmail: Listening for user replies
  - OpenAI: Normalizing free-text replies into structured data
  - Google Calendar: Creating events on confirmation

## Known pitfalls, handle each one explicitly in your implementation
1. Timezone mismatches can cause events to appear at wrong local times, so always convert extracted times to UTC before creating the calendar entry.
2. Personal information leakage occurs if raw emails are logged, so ensure PII is stripped or masked before sending data to the LLM.

## Reference implementation
https://n8n.io/workflows/12167 (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/email-based-scheduling-automation/ via usecasesforagents.com