llms.txt всередині. Промпт введено. Код отримано.

Read this llms.txt:
https://api.paperoffice.ai/latest/docs/llms.txt

Build a small Node.js script that:
1) Uploads an invoice PDF to POST /job with ocr_mode=complete and idp_collection=invoice
2) Uses Authorization: Bearer po_ut_YOUR_API_KEY
3) Sets priority=900 for a synchronous result
4) Prints vendor, total and bounding boxes from job_result

Read this llms.txt:
https://api.paperoffice.ai/latest/docs/llms.txt

Write a curl one-liner and a matching fetch() snippet that:
1) Calls PaperOffice forward geocoding for "Friedrichstraße 123, Berlin"
2) Sends Authorization: Bearer po_ut_YOUR_API_KEY
3) Prints lat, lon and display_name

Read this llms.txt:
https://api.paperoffice.ai/latest/docs/llms.txt

Build a secure document delivery flow:
1) POST /document_delivery/create_and_send with pofids, recipient_email, subject
2) Set security_mode=link (password in the same email) or security_mode=secure_sms
3) Use Authorization: Bearer po_ut_YOUR_API_KEY
4) Print delivery_id and confirm access events are logged

Read this llms.txt:
https://api.paperoffice.ai/latest/docs/llms.txt

Build a Human-in-the-Loop integration that:
1) Creates or fetches an open HITL review task for a processed document
2) Shows extracted fields that need approval
3) Submits approve or reject via the documented HITL endpoints
4) Uses Authorization: Bearer po_ut_YOUR_API_KEY
5) Prints the updated task status

Read this llms.txt:
https://api.paperoffice.ai/latest/docs/llms.txt

Build a document chat (RAG) client that:
1) Asks a question against a workspace or document set
2) Uses Authorization: Bearer po_ut_YOUR_API_KEY
3) Prints the answer plus cited source snippets
4) Handles follow-up questions in the same session when supported

Read this llms.txt:
https://api.paperoffice.ai/latest/docs/llms.txt

Build a Knowledge Base Q&A script that:
1) Sends a user question to the Knowledge Base / helpdesk knowledge endpoints
2) Uses Authorization: Bearer po_ut_YOUR_API_KEY
3) Prints the answer and the article titles used as sources
4) Falls back gracefully when no article matches

Read this llms.txt:
https://api.paperoffice.ai/latest/docs/llms.txt

Build a translation call that:
1) POSTs to /translate/text with source_language=en and target_language=de
2) Sets tone=formal and preserve_terms_preset=paperoffice_brand
3) Translates: "Connect PaperOffice AI-IDP to Claude via MCP without rewriting PaperOffice."
4) Uses Authorization: Bearer po_ut_YOUR_API_KEY
5) Prints the translated text and confirms PaperOffice / AI-IDP / MCP stayed intact

Read this llms.txt:
https://api.paperoffice.ai/latest/docs/llms.txt

Build a TTS flow with pronunciation control:
1) Ensure an account pronunciation lexicon entry for "PaperOffice" (and one custom term)
2) POST /job with text containing those terms, voice=Nadja, output_format=mp3, output=url
3) Set priority=999 for a synchronous audio URL
4) Uses Authorization: Bearer po_ut_YOUR_API_KEY
5) Prints job_result.audio_url

Read this llms.txt:
https://api.paperoffice.ai/latest/docs/llms.txt

Build a PDF splitter that:
1) Uploads a batch PDF via POST /job with template=pdf_ai_split
2) Uses Authorization: Bearer po_ut_YOUR_API_KEY
3) Sets priority high enough for a usable sync/async result per llms.txt
4) Prints the resulting split file list / URLs

Read this llms.txt:
https://api.paperoffice.ai/latest/docs/llms.txt

Build a PII redaction flow that:
1) Detects PII in an uploaded PDF (preview with redact boxes)
2) Applies redaction to produce a GDPR-safe output file
3) Uses Authorization: Bearer po_ut_YOUR_API_KEY
4) Prints detected categories and the redacted file URL / id