You are helping me build the following AI agent workflow.

## Goal
Conversational Telegram: Receives text or voice messages on Telegram, transcribes audio, and generates context-aware AI responses with a typing indicator.

## Specification
- What it does: Receive instant, context-aware replies to your text or voice messages directly inside Telegram without needing to switch apps. The agent transcribes your audio notes and generates thoughtful responses while showing a typing indicator so you know it is working.
- Trigger: Triggered by an event (Event · on new message)
- Autonomy: Fully hands-off
- Expected setup effort: about an afternoon
- Tools/services involved:
  - Telegram: Messaging and triggers
  - OpenAI Chat Model: Text generation
  - Simple Memory: Context management

## Known pitfalls, handle each one explicitly in your implementation
1. Context windows fill up quickly during long chats so implement a sliding window strategy to keep memory usage manageable.

## Reference implementation
https://n8n.io/workflows/4696 (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/conversational-telegram-bot/ via usecasesforagents.com