ocr-expert

ocr-expert is a skill for Claude Code, Codex from sandraschi/ocr-mcp. It costs 0 tokens per session (504 once invoked), scanned A, original, MIT.

A guide for using an OCR server that turns scanned pages and images into text, with 14 possible recognition backends. OCR, or optical character recognition, reads printed or handwritten content from images and documents.

In plain words
What is it for?
Use it to process documents, scan pages, extract tables or structured Markdown, and select between local or cloud OCR engines.
Why use it?
It helps choose an OCR backend suited to the document, language, hardware, and output needs. It also provides fallback options when a preferred backend is unavailable.

Skill for Claude CodeCodex

Part of the ocr-mcp plugin — 4 skills shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/sandraschi/ocr-mcp/ocr-expert
Any agent
npx skills add sandraschi/ocr-mcp --skill ocr-expert
Clone the repo
git clone --depth 1 https://github.com/sandraschi/ocr-mcp

Made for: Claude Code, Codex.

Or install ocr-mcp, the plugin that ships this one along with the rest of its 4 skills.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for ocr-expert

README.md
[![agentmods](https://agentmods.dev/badge/skills/sandraschi/ocr-mcp/ocr-expert.svg)](https://agentmods.dev/skills/sandraschi/ocr-mcp/ocr-expert)
Your own site
<a href="https://agentmods.dev/skills/sandraschi/ocr-mcp/ocr-expert"><img src="https://agentmods.dev/badge/skills/sandraschi/ocr-mcp/ocr-expert.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 504 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.00504
Opus 5 $0.00000 $0.00252
Sonnet 5 $0.00000 $0.00101
Haiku 4.5 $0.00000 $0.00050

Measured 5d ago against content hash ba0b8a3daedb, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ocr-expert scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

src/ocr_mcp/skills/ocr-expert/SKILL.md · 50 lines

What it actually says

OCR Expert Skill

You are connected to OCR-MCP, a FastMCP 3.4+ server with 14 OCR backends. This skill teaches you how to use it effectively.

Available Backends (14)

Backend Params VRAM License Best For
unlimited-ocr 3B ~6GB MIT Long docs, unbounded parsing
paddleocr-vl 0.9B 3.3GB Apache General docs, tables, 109 langs
mineru-2.5 1.2B ~4GB Apache Academic/technical docs
nemotron-vl 8B ~16GB NV Open Invoices, forms, charts (EN only)
deepseek-ocr2 3B ~8GB MIT Structured markdown extraction
olmocr-2 7B ~16GB Apache Academic PDFs, math
mistral-ocr API 0 Cloud High accuracy cloud OCR
got-ocr 580M ~2GB Apache Fast, lean VRAM
tesseract CPU 0 Apache CPU fallback, always available

Default backend: unlimited-ocr. Auto-selection falls back through the chain.

Core Workflows

Single document OCR

process_document(operation="process_document", image_path="doc.png", backend="unlimited-ocr")

Scan then OCR

operate_scanner(operation="scan_document", device_id="...", dpi=300)
process_document(operation="process_document", image_path="<result.path>", backend="unlimited-ocr")

Table extraction

process_document(operation="extract_tables", image_path="table.png", backend="paddleocr-vl")

Key Configuration

  • OCR_DEVICE: cuda (default if GPU), cpu
  • OCR_CACHE_DIR: model download cache
  • MISTRAL_API_KEY: required for mistral-ocr backend
  • OCR_SAMPLING_USE_CLIENT_LLM=1: use host IDE's LLM instead of local Ollama

Agentic Workflow

For multi-step tasks, use execute_agentic_workflow with a natural language goal. The agent will autonomously call tools in sequence.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 5d ago First seen · 50 lines · 0 tokens per session scan A ba0b8a3daedb

Subscribe to this mod's changes

ocr-expert is a skill published in the GitHub repository sandraschi/ocr-mcp (20 stars, last pushed 4d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 504 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

outline

Conventions and efficient workflows for Outline knowledge bases via mcp-outline tools. Use when creating, editing, restructuring, or diagramming Outline documents — covers Outline markdown quirks (mermaidjs code fences), document and collection structure, string-match editing, staged rewrites, batch operations, and…

Vortiago/mcp-outline · 64 tokens

onenote

OneNote MCP connects an AI assistant to Microsoft OneNote through the Microsoft Graph API: browsing notebooks/sections/pages, reading full page content (HTML), creating pages, and full-text search across the user's notes. 13 tools, dual transport (stdio + HTTP streamable on 10907).

sandraschi/onenote-mcp · 0 tokens

work-with-faqs

Installs, updates, and uses FAQ documentation across codebases. Use when working with FAQs, discovering package documentation, updating FAQs after code changes, or when the user asks about FAQ installation, updates, or usage.

fluent-meaning-symbiotic/agentic_executables · 49 tokens

enrich-index

Batch-enrich astrolabe index cards. Use when the user asks to enrich, classify, or describe documents in the astrolabe index, or when listdocs(stale=true) shows unenriched cards. Reads each file, determines its type, writes a summary, keywords, and extracts headings, then calls updateindextool.

zebrr/astrolabe-mcp · 70 tokens

docx

Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when…

tashifkhan/agentic-browser · 168 tokens

xlsx

Use this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or…

tashifkhan/agentic-browser · 201 tokens