adx 0.1.0

ADX: Agentic Data Layer. Structured, citeable, inspectable document state for AI agents.

ADX

Agent-native document intelligence layer. Wraps best-in-class document parsers and exposes structured, citeable, inspectable document state to AI agents.

What It Does

ADX is not a parser. It is an orchestration layer that:

1. Wraps parsers (PyMuPDF, openpyxl, csv) behind a uniform interface
2. Builds a canonical DocumentGraph from any supported format
3. Exposes 9 read-only agent tools for inspection and search
4. Extracts fields using built-in schemas with field-level citations
5. Validates results with rule-based checks

Install

pip install adx

Quick Start

from adx import ADX

dn = ADX()
doc_id = dn.upload("invoice.pdf")

# Profile the document
profile = dn.profile(doc_id)

# Extract with a built-in schema
extraction = dn.extract(doc_id, schema="invoice")

# Validate
result = dn.validate(doc_id, extraction["id"])

Agent Tools

Tool	Description
profile_document	File metadata, type detection, recommended tools
list_structure	Sections, tables, page outline
search_document	Full-text search with citations
get_page	Page text blocks and tables
get_table	Table rows with headers
list_sheets	Spreadsheet sheet metadata
read_range	Cell range with formulas
find_cells	Search cells by value
inspect_formula	Trace formula dependencies

Built-in Schemas

• invoice — vendor, line items, totals, tax
• contract — parties, dates, governing law
• financial_model — revenue, expenses, assumptions
• table — generic table extraction

REST API

adx serve
curl -X POST http://localhost:8000/v1/files -F "file=@invoice.pdf"
curl http://localhost:8000/v1/files/{id}/profile

CLI

adx upload invoice.pdf
adx profile <id>
adx extract <id> --schema invoice
adx validate <id> --extraction <eid>

Supported Formats

Format	Parser	Features
PDF	PyMuPDF	Text blocks, tables, bounding boxes, sections
Excel (.xlsx)	openpyxl	Sheets, formulas, hidden content, named ranges
DOCX	python-docx	Paragraphs, headings, tables, images, metadata
RTF	striprtf	Text extraction with formatting stripped
CSV	stdlib	Dialect sniffing, encoding detection

Documentation

https://harsh-nod.github.io/adx/

License

MIT