You are helping me build the following AI agent workflow.

## Goal
Wikipedia Podcast Generator: Send a topic or voice note to the bot and get back a 5-minute podcast episode researched from Wikipedia and synthesized with AI.

## Specification
- What it does: Send a text or voice note about any topic to instantly receive a five-minute podcast episode researched from Wikipedia. The system transcribes your input, writes a structured script using AI research, and delivers natural-sounding audio directly back to you via Telegram.
- Trigger: Run manually (Event · on new Telegram message)
- Autonomy: Fully hands-off
- Expected setup effort: about an afternoon
- Tools/services involved:
  - Telegram: Receive and send messages
  - Anthropic Chat Model: AI agent processing and script generation
  - OpenAI: Speech transcription via Whisper
  - ElevenLabs API: Text to speech synthesis

## Known pitfalls, handle each one explicitly in your implementation
1. Voice files from Telegram can be large, so implement a download step with size limits to prevent memory overflow during Whisper processing.
2. Wikipedia content may change rapidly, so add a timestamp check in the LLM prompt to prioritize recent edits or stable versions for accuracy.
3. ElevenLabs API costs scale with character count, so cap the script length to ensure the final audio stays within the five-minute target duration.
4. Telegram has strict file size limits for outgoing media, so compress the generated MP3 or split long responses if they exceed 50MB.

## Reference implementation
https://n8n.io/workflows/4496 (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/wikipedia-podcast-generator/ via usecasesforagents.com