Borrowing it
Nothing to install: this file belongs to tinygone/unlimited-ocr-mcp-server. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/tinygone/unlimited-ocr-mcp-server/master/CLAUDE.mdgit clone --depth 1 https://github.com/tinygone/unlimited-ocr-mcp-serverWrote 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.
[](https://agentmods.dev/instructions/tinygone/unlimited-ocr-mcp-server/claude-md)<a href="https://agentmods.dev/instructions/tinygone/unlimited-ocr-mcp-server/claude-md"><img src="https://agentmods.dev/badge/instructions/tinygone/unlimited-ocr-mcp-server/claude-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.01684 | $0.01684 |
| Opus 5 | $0.00842 | $0.00842 |
| Sonnet 5 | $0.00337 | $0.00337 |
| Haiku 4.5 | $0.00168 | $0.00168 |
Grade A, and why
unlimited-ocr-mcp-server CLAUDE.md scanned grade A with 1 finding 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl http://127.0.0.1:10000/v1/models How it starts
The opening of the file, as written. The whole thing — 90 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
项目概述
把百度的 Unlimited-OCR 模型封装成 MCP 工具,让 Claude Code 能识别任意图片中的文字。本地推理、无调用配额、无 API 费用。整套链路必须在本机 GPU 上运行。 我在Windows 11 + RTX5070TI LAPTOP,使用Python 3.12.3,CUDA13.2,测试成功。 官网推荐是Python3.12.3 + CUDA12.9。如果使用其他版本,建议看看测试效果。
常用命令
服务必须在 conda 环境 uocr 中运行(Python 3.12.3 + PyTorch/CUDA)。
conda activate uocr
# 1. 启动 OCR 服务(关键,必须常驻。首次下载模型需设镜像)
$env:HF_ENDPOINT = "https://hf-mirror.com" # PowerShell; bash 用 export
$env:HF_HOME = "your/path/hf/cache" # PowerShell; bash 用 export。换成你自己的缓存目录(如 D:/hf/cache)
python serve_ocr.py
# 2. 验证服务在线(MCP 能用的前提)
curl http://127.0.0.1:10000/v1/models
# 3. 测试 HTTP API(默认图 inputs/test.jpg; --image 指定图; --stream 测流式)
python test_api.py --image inputs/1.jpg
# 4. 命令行直接推理(脱离 MCP,单进程内加载模型并跑一次)
python run_ocr.py --mode single -i inputs/1.jpg -o outputs
python run_ocr.py --mode multi -i inputs/p1.png inputs/p2.png -o outputs
python run_ocr.py --mode pdf -i inputs/doc.pdf -o outputs --dpi 300
# 5. 调试脚本:探测 model.infer 的真实返回值类型
python probe_infer.py
MCP server(ocr_mcp_server.py)无需手动启动——Claude Code 通过 .mcp.json 自动拉起。
本项目没有自动化测试框架。test_api.py / probe_infer.py 是手动 smoke-test 脚本,不是 pytest 用例。
架构:两层解耦(改代码前必读)
Claude Code ──MCP(stdio)──> ocr_mcp_server.py ──HTTP /v1/chat/completions──> serve_ocr.py ──> 模型
| 层 | 文件 | 职责 |
|---|---|---|
| 模型服务 | serve_ocr.py |
启动时一次性加载模型进显存,FastAPI 暴露 OpenAI 兼容 HTTP API(:10000),常驻进程 |
| MCP 网关 | ocr_mcp_server.py |
把 HTTP API 包装成 MCP 工具 ocr_image / ocr_multi,被 Claude Code 经 stdio 调用 |
为什么分两层:模型加载耗时数十~上百秒。拆开后模型只在 serve_ocr.py 启动时加载一次,Claude Code / MCP server 重启都不重载模型。这是整个设计的核心不变量,改动时不能破坏。
两个推理分支(参数不同,别混用)
serve_ocr.py 的 run_inference() 按图片数量走不同模型方法:
- 单图 →
model.infer(),走 gundam 配置(base_size=1024, image_size=640, crop_mode=True),适合大图切片 - 多图 →
model.infer_multi()(image_size=1024),保留跨页上下文
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.
- 5d ago First seen · 90 lines · 1,684 tokens per session scan A 8344eab0ddaf
unlimited-ocr-mcp-server CLAUDE.md is an instructions file published in the GitHub repository tinygone/unlimited-ocr-mcp-server (0 stars, last pushed 1mo ago), licensed MIT. It adds 1,684 tokens to every session, about $0.0084 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.