Skip to content

reductoai/claude-plugins

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Reducto Claude Code Plugins

Parse, extract, and edit PDFs, images, spreadsheets, and Office documents directly inside Claude Code — powered by Reducto's document intelligence API.

Reducto turns unstructured documents into structured, LLM-ready data using agentic OCR and vision-language models. This plugin brings those capabilities into your Claude Code workflow so you can process invoices, contracts, medical records, financial filings, and more without leaving your terminal.

Installation

Via Vercel Skills CLI (recommended)

npx skills add reductoai/claude-plugins

Via Claude Code Plugin System

Add this marketplace to Claude Code:

/plugin marketplace add reductoai/claude-plugins

Then install the plugin:

/plugin install reducto-cli@reducto-plugins

Quick Start

# 1. Authenticate with your Reducto account
uvx --from reducto-cli reducto login

# 2. Parse a PDF into clean Markdown
uvx --from reducto-cli reducto parse document.pdf

# 3. Extract structured JSON from an invoice
uvx --from reducto-cli reducto extract invoice.pdf --schema schema.json

# 4. Fill out a form with natural language
uvx --from reducto-cli reducto edit form.pdf --instructions "Fill in name as 'John Doe'"

Plugin: reducto-cli

Parse — Convert Documents to Markdown

Transform any supported document into clean, structured Markdown with optional YAML metadata. Parsed output preserves document layout, tables, figures, and semantic structure.

uvx --from reducto-cli reducto parse document.pdf
uvx --from reducto-cli reducto parse ./documents/   # batch process a directory

Options:

Flag Description
--agentic Maximum accuracy mode — uses agentic OCR for tables, text, and figures (slower)
--change-tracking Outputs <s>, <u>, and <change> tags for tracked revisions
--highlights Includes highlighted text in output
--hyperlinks Preserves embedded hyperlinks
--comments Includes document comments and annotations
# High-accuracy parse with full metadata
uvx --from reducto-cli reducto parse contract.pdf --agentic --change-tracking --comments

Output files are written as <filename>.parse.md.

Extract — Pull Structured Data with JSON Schema

Define the data you need using a JSON Schema and Reducto extracts it into typed JSON. The top-level schema type must be object.

uvx --from reducto-cli reducto extract invoice.pdf --schema schema.json

Example schema for invoice extraction:

{
  "type": "object",
  "properties": {
    "invoice_number": { "type": "string" },
    "date": { "type": "string" },
    "vendor": {
      "type": "object",
      "properties": {
        "name": { "type": "string" },
        "address": { "type": "string" }
      }
    },
    "line_items": {
      "type": "array",
      "items": {
        "type": "object",
        "properties": {
          "description": { "type": "string" },
          "quantity": { "type": "number" },
          "unit_price": { "type": "number" },
          "total": { "type": "number" }
        },
        "required": ["description", "quantity", "unit_price", "total"]
      }
    },
    "total": { "type": "number" }
  },
  "required": ["invoice_number", "line_items", "total"]
}

You can also pass schemas inline:

uvx --from reducto-cli reducto extract receipt.pdf --schema '{"type":"object","properties":{"total":{"type":"number"},"date":{"type":"string"}}}'

Output files are written as <filename>.extract.json.

Edit — Modify Documents with Natural Language

Programmatically fill forms, update fields, and modify documents using plain English instructions. Edited files are returned in their original format.

uvx --from reducto-cli reducto edit application.pdf -i "Fill out: Name: Jane Smith, Email: jane@example.com"
uvx --from reducto-cli reducto edit contract.pdf -i "Set the effective date to March 1, 2025 and the client name to Acme Corp"
uvx --from reducto-cli reducto edit ./forms/ -i "Check the 'Approved' box and sign with today's date"

Output files are written as <filename>.edited.<ext>.

Supported File Types

Category Formats
PDF .pdf
Images .png, .jpg, .jpeg
Word Documents .doc, .docx
Presentations .ppt, .pptx
Spreadsheets .xls, .xlsx

Reducto handles scanned documents, faxes, and handwritten content through its OCR engine, and supports 100+ languages including mixed-language documents.

Use Cases

Invoice and Receipt Processing

Parse a folder of invoices, extract line items, totals, and vendor info into JSON, then feed the structured data into your accounting pipeline.

uvx --from reducto-cli reducto extract ./invoices/ --schema invoice_schema.json

Contract Review and Clause Extraction

Convert legal contracts to Markdown with change tracking enabled, then extract specific clauses, dates, and party names.

uvx --from reducto-cli reducto parse contract.pdf --agentic --change-tracking
uvx --from reducto-cli reducto extract contract.pdf --schema '{"type":"object","properties":{"effective_date":{"type":"string"},"termination_clause":{"type":"string"},"parties":{"type":"array","items":{"type":"string"}}}}'

Financial Document Analysis

Extract tables and figures from SEC filings, earnings reports, and investor decks with high-accuracy agentic parsing.

uvx --from reducto-cli reducto parse 10-K_filing.pdf --agentic

Automated Form Filling

Batch-fill PDF application forms, onboarding packets, or compliance documents using natural language instructions.

uvx --from reducto-cli reducto edit ./onboarding_forms/ -i "Fill in employee name: Alex Chen, start date: 2025-04-01, department: Engineering"

RAG Pipeline Document Preparation

Parse documents into structured Markdown optimized for chunking and embedding in retrieval-augmented generation systems.

uvx --from reducto-cli reducto parse ./knowledge_base/ --agentic

Medical Record Summarization

Extract patient info, diagnoses, and treatment plans from medical documents using a custom schema.

Insurance Claims Processing

Parse claim forms and supporting documents, then extract policy numbers, incident details, and damage assessments into structured data.

Multi-Language Document Processing

Process documents in 100+ languages — Reducto handles mixed-language content and non-Latin scripts through its vision-language models.

Requirements

  • Claude Code 1.0.33 or later
  • UV (automatically installed by the plugin if not present)
  • A Reducto account — sign up at studio.reducto.ai

Resources

License

MIT

About

Official Reducto plugins for Claude Code

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors