Dnvriend Ollama Deepseek Ocr Tool
dnvriend-ollama-deepseek-ocr-tool is a Claude Code plugin for the Build phase that exposes a CLI to batch OCR document images with DeepSeek-OCR through Ollama.
by dnvriend · github.com/dnvriend/ollama-deepseek-ocr-tool
Batch-convert document images to markdown text locally by invoking DeepSeek-OCR through Ollama from Claude Code or your terminal.
Add it to Claude Code
Install the plugin in Claude Code. One command, paste-ready.
/plugin install dnvriend-ollama-deepseek-ocr-tool@dnvriend/ollama-deepseek-ocr-toolBuilt to be called by your agent
Skillselion is itself an MCP server. Your agent can pull this entry and a paste-ready install config straight from the API - no copy-paste.
Retrieve this entry with skillselion.get_details("plugin:dnvriend/ollama-deepseek-ocr-tool") and the paste-ready config with skillselion.get_install_config("plugin:dnvriend/ollama-deepseek-ocr-tool").
What it does
dnvriend-ollama-deepseek-ocr-tool is a Claude Code plugin wrapping one CLI utility that batch-processes document images into markdown using DeepSeek-OCR served by Ollama on your machine. Solo builders digitizing receipts, scans, screenshots, or PDF exports for RAG, support macros, or content pipelines install it when they want OCR without shipping pixels to a proprietary cloud API. The tool loads image paths, applies an OCR-oriented prompt, and returns structured text the agent can chunk, summarize, or load into a database. It is phase-specific integration work: you run it during build when wiring ingestion, or occasionally in operate when reprocessing archived scans—not for brainstorming or SEO. Complexity is beginner on the happy path if Ollama and the model are already installed, but intermediate when batch sizes, GPU memory, or path automation matter. This is a narrow task tool, not a journey-wide methodology skill.
Highlights
- Single-plugin CLI for batch OCR on document images via Ollama.
- Uses DeepSeek-OCR model with instruct-style prompts and markdown output.
- Local-first: no cloud OCR API required when Ollama is running.
- Designed as an invokable tool from agent sessions (loads images, returns text).
- Community plugin (dnvriend/ollama-deepseek-ocr-tool) with focused scope.
Why builders use it
Builders with local LLM stacks still retype or manually copy text from scanned documents because they lack a simple batch OCR bridge for agents.
After install, you can invoke batch image-to-markdown OCR through Ollama and feed the text straight into downstream build or content workflows.
At a glance
- Type - Plugin in Automation.
- Adoption - 0 installs, 3 stars, 0 votes.
FAQ
Who is dnvriend-ollama-deepseek-ocr-tool for?
It is for developers and solo builders using Claude Code who run Ollama locally and want DeepSeek-OCR for document and image batches.
When should I use dnvriend-ollama-deepseek-ocr-tool?
Use it when ingesting scanned or image-based documents into markdown during build or content automation—not for git worktrees or launch SEO.
How do I add dnvriend-ollama-deepseek-ocr-tool to my agent?
Install the plugin from dnvriend/ollama-deepseek-ocr-tool in Claude Code, install Ollama with the DeepSeek-OCR model, then invoke the CLI tool with your image paths from the agent or shell.
Comments
Share how you use dnvriend-ollama-deepseek-ocr-tool, gotchas, or tips for other indie builders.
No comments yet - be the first to share how you use it.