Borrowing it
Nothing to install: this file belongs to magicyuan876/mineru-tianshu. 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/magicyuan876/mineru-tianshu/main/AGENTS.mdgit clone --depth 1 https://github.com/magicyuan876/mineru-tianshuWrote 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/magicyuan876/mineru-tianshu/agents-md)<a href="https://agentmods.dev/instructions/magicyuan876/mineru-tianshu/agents-md"><img src="https://agentmods.dev/badge/instructions/magicyuan876/mineru-tianshu/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/magicyuan876/mineru-tianshu/agents-md"><img src="https://agentmods.dev/badge/instructions/magicyuan876/mineru-tianshu/agents-md.svg" alt="Reviewed on agentmods" width="80" 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 | $0.04665 | $0.04665 |
| Opus 5 | $0.02332 | $0.02332 |
| Sonnet 5 | $0.00933 | $0.00933 |
| Haiku 4.5 | $0.00466 | $0.00466 |
Grade A, and why
mineru-tianshu AGENTS.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 today.
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.
功能验证方式：`make test-gpu`（容器内验证 CUDA）、`make test-api`（curl 健康检查），以及通过前端页面/API 手动跑通端到端流程。 How it starts
The opening of the file, as written. The whole thing — 253 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — MinerU Tianshu（天枢）
本文件面向 AI 编码代理，介绍本项目的架构、命令与开发约定。 项目文档、代码注释、日志输出以简体中文为主，修改现有文件时请保持中文注释风格。
项目概述
MinerU Tianshu（天枢）是一个企业级 AI 数据预处理平台，将非结构化数据转换为 AI 可用的结构化格式（Markdown + JSON）：
- 📄 文档：PDF、Word、Excel、PPT（MinerU 原生解析 DOCX/XLSX/PPTX；旧版 .doc/.xls/.ppt 经 LibreOffice 转换）、HTML/TXT/CSV（MarkItDown）
- 🖼️ 图片：JPG、PNG、BMP、TIFF（多 OCR 引擎 + 水印去除🧪）
- 🎙️ 音频：MP3、WAV、M4A、FLAC（SenseVoice 多语言、说话人识别、情感识别）
- 🎬 视频：MP4、AVI、MKV、MOV、WebM（FFmpeg 音频提取转写 + 关键帧 OCR🧪）
- 🧬 生物格式：FASTA、GenBank（插件化格式引擎）
- 🏗️ 企业特性：GPU 负载均衡、任务队列、JWT 认证、API Key、MCP 协议、RustFS 对象存储
技术栈
| 层 | 技术 |
|---|---|
| 前端 | Vue 3、TypeScript、Vite、TailwindCSS、Pinia、Vue Router、vue-i18n（中英文） |
| 后端 | Python 3.12+、FastAPI、LitServe（GPU 负载均衡）、SQLite（任务队列/认证）、可选 Redis 队列 |
| AI 引擎 | MinerU 3.4+（pipeline/vlm/hybrid，含 vllm-mineru 远程模式）、SenseVoice/FunASR、MarkItDown |
| 基础设施 | Docker Compose、Nginx（前端）、RustFS（S3 兼容对象存储，minio-py 客户端）、MCP 协议 |
| 许可证 | Apache License 2.0 |
项目结构
mineru-tianshu/
├── frontend/ # Vue 3 前端
│ ├── src/
│ │ ├── api/ # Axios API 客户端（authApi/taskApi/queueApi/systemApi）
│ │ ├── components/ # 组件（MarkdownViewer、JsonViewer、VirtualPdfViewer 等）
│ │ ├── views/ # 页面（Dashboard、TaskSubmit、TaskDetail、Login 等）
│ │ ├── stores/ # Pinia 状态（authStore、taskStore、queueStore 等）
│ │ ├── locales/ # i18n（zh-CN.ts、en-US.ts）
│ │ ├── router/ # Vue Router（含权限路由守卫）
│ │ └── types/ utils/ # TypeScript 类型与工具
│ ├── package.json vite.config.ts Dockerfile
│
├── backend/ # Python 后端（所有服务共用此目录代码）
│ ├── start_all.py # 统一启动脚本（本地开发入口）
│ ├── api_server.py # FastAPI API 服务器（端口 8000）
│ ├── litserve_worker.py # LitServe GPU Worker Pool（端口 8001）
│ ├── task_scheduler.py # 任务调度器（超时重置、文件清理）
│ ├── mcp_server.py # MCP 协议服务器（端口 8002，供 Claude Desktop 等调用）
│ ├── task_db.py # SQLite 任务数据库（启动时原地迁移）
│ ├── redis_queue.py # Redis 优先级队列（可选，SQLite 自动回退）
│ ├── download_models.py # 模型预下载脚本
│ ├── auth/ # JWT 认证、API Key、角色权限、OIDC/SAML SSO
│ ├── mineru_pipeline/ # MinerU 引擎封装
│ ├── audio_engines/ # SenseVoice 音频引擎
│ ├── video_engines/ # 视频引擎（FFmpeg + 关键帧提取，OCR 走 MinerU）
│ ├── format_engines/ # 插件化格式引擎（FASTA、GenBank 为参考实现）
│ ├── remove_watermark/ # 水印去除（YOLO11x + LaMa）
│ ├── output_normalizer/ # 输出标准化（统一 result.md/result.json/images/）
│ ├── storage/ # RustFS S3 客户端（图片上传、URL 替换）
│ ├── utils/ # 工具函数
│ ├── requirements.txt # Python 依赖（版本锁定，含 mineru[all]>=3.4.5）
│ └── Dockerfile Dockerfile.cpu Dockerfile.offline install.sh
│
├── dify_plugin/tianshu/ # Dify 插件（manifest.yaml + provider/ + tools/）
├── setup.sh # 一键部署入口（交互式引导 + 非交互参数 --mode/--yes/--dry-run）
├── scripts/ # 辅助脚本（docker-entrypoint.sh、init-models.sh）
├── docker-compose.yml # 生产编排（GPU 版，含 vllm-mineru/rustfs 等）
├── docker-compose.dev.yml # 开发编排（热重载 + debugpy 端口 5678）
├── docker-compose.cpu.yml # CPU 本地开发（Mac Apple Silicon，需 Rosetta 2）
├── docker-compose.offline.yml # 离线部署编排
├── docker-compose.pipeline.yml# 纯 pipeline 模式部署
├── Makefile # Docker 快捷命令
├── pyproject.toml # Ruff 配置（无 Python 打包用途）
├── .pre-commit-config.yaml # pre-commit 钩子（唯一 CI 质量门禁）
└── .env.example # 根环境变量模板（docker-compose 与 Makefile 消费）
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.
- today First seen · 253 lines · 4,665 tokens per session scan A e1512fe78d01
mineru-tianshu AGENTS.md is an instructions file published in the GitHub repository magicyuan876/mineru-tianshu (797 stars, last pushed yesterday), licensed Apache-2.0. It adds 4,665 tokens to every session, about $0.0233 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 2026-09-12.
Other instructions, from other repositories
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.
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).
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.
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).