llm_vision: Instructions file for Claude Code

CLAUDE.md

llm_vision CLAUDE.md is an instructions file for Claude Code from 1710782766/llm_vision. It costs 1,734 tokens per session, scanned A, original, MIT.

A set of project instructions for Claude Code, an AI coding assistant, when working on a local MCP server that lets models analyze images and read text from them.

In plain words
What is it for?
Use it to install dependencies, run tests, start the local server, analyze images, extract text, and compare vision models.
Why use it?
It gives the assistant the project's structure, commands, setup details, and image-handling rules so it can work with the repository more accurately.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: mentions CLAUDE.md; mentions Claude Code.

This is 1710782766/llm_vision's own configuration. It tells Claude Code how to work on llm_vision itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything llm_vision configures →

Reuse

Borrowing it

Nothing to install: this file belongs to 1710782766/llm_vision. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/1710782766/llm_vision/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/1710782766/llm_vision

Made for: Claude Code.

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 llm_vision CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/1710782766/llm_vision/claude-md/github.svg)](https://agentmods.dev/instructions/1710782766/llm_vision/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/1710782766/llm_vision/claude-md"><img src="https://agentmods.dev/badge/instructions/1710782766/llm_vision/claude-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.

agentmods 80×15 button for llm_vision CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/1710782766/llm_vision/claude-md"><img src="https://agentmods.dev/badge/instructions/1710782766/llm_vision/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 1,734 This file is loaded in full into every session.
When invoked 1,734 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.01734 $0.01734
Opus 5 $0.00867 $0.00867
Sonnet 5 $0.00347 $0.00347
Haiku 4.5 $0.00173 $0.00173

Measured 8d ago against content hash 3784f9092027, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

llm_vision CLAUDE.md 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 8d 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.

CLAUDE.md · 51 lines

How it starts

The opening of the file, as written. The whole thing — 51 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

本地 MCP 服务器:为无视觉能力的模型(如 deepseek 系)提供视觉——模型传入本地图片路径,服务器调用阿里 DashScope 视觉模型,返回文字结果。已注册在 .mcp.json(可移植形式,cwd: "."),本项目内任何 Claude Code 会话直接可用 describe_image / extract_textimages/ 目录被 gitignore(测试图不入库),冒烟/对比脚本可用参数传入图片路径或自备 images/ 测试图。

常用命令

uv sync                                     # 安装依赖(Python 3.12,uv 管理)
uv run pytest tests/ -q                     # 全量测试(68 个,多数离线不花钱)
uv run pytest tests/test_dashscope_client.py -v   # 单文件测试
uv run python main.py                       # 启动 MCP 服务器(stdio;Claude Code 会按 .mcp.json 自动拉起)
uv run python scripts/smoke_test.py [图片路径 ...]   # 真实 API 冒烟(消耗额度;绕缓存;缺省用 images/ 自备图,无图时提示传路径)
uv run python scripts/compare_models.py qwen3-vl-plus qwen3.7-plus   # 模型对比(同样耗额度)

架构

main.py(FastMCP 薄注册)
  ├── describe_image(image_path, prompt?, perspective="normal")   → LLM_VISION_MODEL(默认 qwen3-vl-plus)
  ├── extract_text(image_path, prompt?)     → LLM_VISION_OCR_MODEL(默认 qwen3.5-ocr)
  └── 共用 _analyze_image 管线
        → image_cache 命中检查(可选注入,内容寻址,仅存文本回答)
        → image_loader.load_image(路径/扩展名白名单/10MB 校验,base64+MIME 推断;HEIC/HEIF 走 macOS sips 转换)
        → image_preprocess.preprocess_image(超 1568px 等比缩放 / 超 1.5MB 重压,macOS sips 零依赖,临时文件处理后即删;无 sips 平台跳过)
        → dashscope_client.DashscopeClient.chat(超时/网络/5xx 重试,等比递减预算 ≤2×timeout)
  • 工具恒返回字符串:成功返回模型回答,失败返回可读中文错误信息,绝不抛异常到客户端。错误消息以「文件不存在」「不支持的图片格式」「图片过大」「视觉模型调用失败」等开头——测试断言依赖这些前缀,改动错误文案需同步测试
  • 依赖注入DashscopeClient(api_key, timeout, max_retries=2, transport=None)transport 仅测试注入(httpx.MockTransport),生产传 None。测试注入模式:max_retries=0(即时失败,不等待退避)、缓存测试用 tmp_path 作 cache_dir、重试测试 monkeypatch.setattr("time.sleep", ...) 跳过退避。新增测试不要发真实请求
  • mcp SDK 是 2.0.0mcp.server.fastmcp 已移除,用 from mcp.server import MCPServer as FastMCP;断言工具注册可用 mcp._tool_manager(私有接口),端到端验证用 stdio 子进程(tests/test_main.py 里有完整范式)
  • 模型可配置:环境变量 LLM_VISION_MODEL / LLM_VISION_OCR_MODEL / LLM_VISION_TIMEOUT(默认 120)/ LLM_VISION_MAX_RETRIES(默认 2)/ LLM_VISION_DESCRIBE_PROMPT / LLM_VISION_OCR_PROMPT / LLM_VISION_CACHE(默认开)/ LLM_VISION_CACHE_DIR / LLM_VISION_MAX_EDGE(默认 1568)/ LLM_VISION_COMPRESS(默认开)。注意 qwen3-vl-plus-latest 别名已 404 失效,勿用
  • 改默认值需同步 5 处config.py 的 DEFAULT_* 常量、tests/test_config.pytest_defaults 断言、README ×2 环境变量表、CLAUDE.md 本条、docs/model-eval-2026-08-05.md 结论——只改常量不同步测试会静默失败(此前踩过)
  • _analyze_image 签名约定:第 5、6 参数 max_edge / compress_enabled(预处理),第 7 参数 cache(ImageCache | None)。cache 默认 None = 不缓存——smoke 脚本天然绕过缓存验证真实 API;错误字符串永不入缓存
  • 升级模型前先读 docs/model-eval-2026-08-05.md:三模型对比结论(qwen3-vl-plus 唯一零幻觉、qwen3.7-plus 为候选、qwen3.5-omni-plus 有视觉错误),选新模型以它为依据
  • 工具 docstring 即 MCP 工具描述:模型靠它理解工具意图,修改 docstring 会改变模型调用行为
  • describe_image 双视角perspective 参数二选一——normal(默认,DEFAULT_NORMAL_DESCRIBE_PROMPT 自然描述)与 criticalDEFAULT_DESCRIBE_PROMPT 审视视角:客观描述 + 主动报告异常 + 区分事实/推测 + 不找补)。docstring 明确写了「用户提及页面/界面问题(不好看/有问题/检查/找 bug)必须用 critical」——视觉模型会把渲染 bug 合理化,审视指令是实测教训(2026-08-06)。显式 prompt 优先于视角默认;LLM_VISION_DESCRIBE_PROMPT 只覆盖 critical 视角。改这两段提示词会改变模型输出风格,评估文档/README 如引用旧文案需同步

Read the full file on GitHub · 51 lines

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. 8d ago First seen · 51 lines · 1,734 tokens per session scan A 3784f9092027

Subscribe to this mod's changes

llm_vision CLAUDE.md is an instructions file published in the GitHub repository 1710782766/llm_vision (0 stars, last pushed 1mo ago), licensed MIT. It adds 1,734 tokens to every session, about $0.0087 per session on Opus 5. 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 2026-08-31.

Related

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.

vercel/next.js · 7,296 tokens

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.

openai/codex · 5,153 tokens

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).

microsoft/vscode · 6,785 tokens

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).

microsoft/vscode · 5,001 tokens

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.

langchain-ai/langchain · 4,469 tokens

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.

github/spec-kit · 7,104 tokens