verbatim-input-transcribed-from-tg-exchange

---
name: verbatim-input-transcribed-from-tg-exchange
description: Extract the verbatim transcribed voice input text for a Telegram exchange by first calling extract-tg-exchange, then finding the transcription in session transcripts
---

# verbatim-input-transcribed-from-tg-exchange — Get the verbatim transcribed user input for a Telegram exchange

## Arguments

Same context-based input as `extract-tg-exchange`:
- `/verbatim-input-transcribed-from-tg-exchange the one where I asked about logging`
- `/verbatim-input-transcribed-from-tg-exchange message 2212`
- `/verbatim-input-transcribed-from-tg-exchange the rule 4 exchange, save to my_input.md`

## Behavior

1. **Call `/extract-tg-exchange`** with the user's context to identify the specific Telegram exchange and get its metadata.

2. **From the metadata**, extract the `attachment_file_id` (for voice messages) or the text content (for text messages).

3. **For voice messages**: use the file_id and timestamp to locate the corresponding transcription in the current session's conversation context. The transcription was produced by `transcribe.py` during the session. Search the conversation history for the transcription output that matches this exchange.

4. **Return**:
   - The exchange metadata (raw channel block)
   - The verbatim transcribed input text string

5. **If the user requests saving to a file**, write both the metadata and the verbatim transcribed text to that file.

## Output Format

Present with labeled sections:
- **Metadata** (the raw channel block)
- **Verbatim Transcribed Input** (the exact transcription text)

## Notes

- For text messages (not voice), the "verbatim input" is simply the text content from the channel tag — no transcription needed.
- If the transcription cannot be found in the current session context, inform the user that the transcription is not available in this session and suggest re-transcribing the audio file if the .oga file is still in the inbox.