flexorch-mcp

Convert unstructured business documents (PDF, DOCX, invoices, contracts, payroll sheets) into structured, LLM-ready datasets with automatic classification…

От сообществаРаботаетНужен API-ключГлобальныйБесплатноТолько чтение

Что умеет

  • Document.Process: Submit a document for processing — Step 1 of 5. Returns job_id immediately; poll job.status until completed.
  • Job.Status: Poll a job until it finishes — Step 2. Call every 3-5s until status is completed or failed.
  • Job.Result: Read structured fields extracted from a completed document — Step 3.

Какие данные видит

Нужен ли аккаунт

Нужен API-ключ из настроек сервиса

Convert unstructured business documents (PDF, DOCX, invoices, contracts, payroll sheets) into structured, LLM-ready datasets with automatic classification, field extraction, PII masking, and quality scoring.

6 async tools:

  • process_document — submit document for processing
  • get_job_status — poll until complete
  • get_extraction_result — read structured fields + quality grade
  • build_dataset — package records into a named dataset
  • export_dataset — download as JSONL, RAG, CSV, or Markdown
  • search_documents — keyword + semantic search across datasets

Requires: FlexOrch API key — get one at app.flexorch.com/settings

Список инструментов сервера (8)

Технические названия из tools/list. Нужны только разработчикам.

document.processSubmit a document for processing — Step 1 of 5. Returns job_id immediately; poll job.status until completed.
job.statusPoll a job until it finishes — Step 2. Call every 3-5s until status is completed or failed.
job.resultRead structured fields extracted from a completed document — Step 3.
dataset.buildPackage extracted records into a dataset for export — Step 4. Returns job_id; poll job.status until completed.
dataset.searchSearch across all indexed FlexOrch datasets by keyword or semantic meaning.
dataset.exportDownload all records from a built dataset as text — Step 5, final step.
dataset.indexTrigger semantic indexing for a dataset (Pro+ plan required). Must be called before dataset.chunks. Indexing is idempotent and typically completes in 10–60 seconds.
dataset.chunksRetrieve LangChain/LlamaIndex-ready text chunks from an indexed dataset (Pro+ plan required). Dataset must be indexed first via dataset.index.