Visual Memory Translator logo

Visual Memory Translator

Community
TanShilongMario
visual-memory-translator

Reinterprets photos into editorial, artist-book memory pages, and can also translate a sentence or concept into a sparse editorial metaphor card. Not a filter, style-transfer, quote-on-photo template, or full portrait redraw. Analyzes what should survive, selects display/layout/style, and builds image prompts with high whitespace and controlled abstraction. When a photo is given and no style is named, first creates a 4/6/9-panel style-preview contact sheet, then regenerates the chosen direction from the original photo. Includes layered white-border stickers and rounded bold-monoline blocks. Use when the user invokes 影像转译, Visual Memory Translator, 视觉记忆转译, 文本转译, editorial photo reinterpretation, a concept/quote visual card, a holiday-limited editorial overlay, 纸币样张, 纸币实景, 中央邮票, 黑胶唱片实景, or asks to turn a photo into a memory page. Around Qixi, Christmas, and other CN/JP/US holidays (±1 day), may add a sparse seasonal motif unless the user opts out.

Overview

PublisherTanShilongMario
Repositoryvisual-memory-translator-SKILL
Skill namevisual-memory-translator
Stars
189
Forks
18
Bundled files
13
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 13 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by TanShilongMario on GitHub. Read the source before you install it.

Installation

Install the Visual Memory Translator AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/TanShilongMario/visual-memory-translator-SKILL/tree/main/visual-memory-translator
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/TanShilongMario/visual-memory-translator-SKILL.git /tmp/visual-memory-translator-SKILL
mkdir -p .claude/skills
cp -r /tmp/visual-memory-translator-SKILL/visual-memory-translator .claude/skills/visual-memory-translator
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Visual Memory Translator in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Visual Memory Translator on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Visual Memory Translator is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Visual Memory Translator / 影像转译编辑器

Version: 1.5.0 Core principle: 原图是现实记录,新图是记忆转译。无图时,句子是概念,画面是隐喻。

将用户照片转译为具有当代编辑设计、艺术出版、视觉手札气质的二次创作图。

不是:滤镜、风格迁移、普通拼贴、把照片完整重画成肖像、金句叠照片。

是:理解输入 → 决定该留下什么 → 重构版式与隐喻 → 留白 → 输出新作品。照片走记忆转译;句子走概念转译。


When to use

  • 用户说:启用影像转译 / 纸币样张 / 纸币实景 / 中央邮票 / 黑胶唱片实景 / @VisualMemoryTranslator / /visual-memory-translator
  • 用户上传照片并要求做成艺术出版页、记忆页、展览票、邮票记忆等编辑设计感图像
  • 用户只给一句文案、金句、概念,要求做成编辑感视觉卡 / 文本转译
  • 用户指定自选模板(纸币样张、纸币实景、中央邮票、黑胶唱片实景等):跳过预览,直达对应构图
  • 需要输出:视觉 prompt + 图像生成/编辑指令

Input routing

输入路径
有参考图input_mode: photo — 影像记忆转译(主路径)
无图,只有句子/概念/标题input_mode: text — 文本隐喻转译,不要先追问必须上传图片
图 + 指定文案input_mode: mixed — 图走记忆转译,文案作画面文字,不另发明金句

无图且也没有可转译的句子时,再要求上传图片或给出句子。


Core philosophy

  1. Reality as evidence — 原图是现实证据
  2. Translation as memory — 转译承担记忆 / 情绪 / 残影
  3. Less, but precise — 元素少,但每个都有理由
  4. Whitespace is content — 留白是作品本身
  5. Editorial before decorative — 版式优先于装饰
  6. Interpret, do not trace — 概括重组,禁止机械临摹
  7. Style may change, aesthetic discipline must not — 风格可变,极简克制不变

最终应像:有人记住这一刻,编辑了记忆,并设计成一页。
不应像:AI 又生成了这张照片的另一个版本。


Global aesthetic (always on)

要有:极简、克制、安静、呼吸感、大留白、非对称优先、焦点明确、艺术出版 / Editorial / Artist Book 气质、主题色统一。

避免:电商/旅游宣传、模板拼贴、满堂元素、花哨贴纸边框、大量文字、鸡汤、过度复古做旧、廉价手账、转译铺满、逐元素复制原图、无故堆无关物体、金句叠夜景、完整肖像重画。

默认背景:暖米白艺术纸、低纹理、无做旧。


Workflow

Copy and track:

- [ ] 1. 判定 input_mode:photo / text / mixed
- [ ] 2. 读取用户参数;检测日期是否落在节日 ±1 天
- [ ] 3. photo 路径判断是否需要风格预览
- [ ] 4. photo 且未指定风格:先生成一张 4/6/9 格预览图并等待选择
- [ ] 5. text 路径:提炼一个隐喻,默认直出编辑卡(见 text-visual.md)
- [ ] 6. 决定 DISPLAY / LAYOUT / STYLE / 抽象度 / 节日层
- [ ] 7. 构造成品 prompt(photo 必须从原始照片重建)
- [ ] 8. 用质量清单自检;失败则按 recovery 修正重试

Interaction

  1. 读取明确参数 → 分析输入;能合理默认则不追问。
  2. photo 且未指定 style_mode:默认生成一张 6 格(2×3)预览图,而不是直接出成品。
  3. 用户可要求 4 格(2×2)或 9 格(3×3);未说数量时用 6 格。
  4. 预览图内只标 01–09,不把风格名与长说明塞进图里;在回复中逐号列出风格名和一句适配理由。
  5. 用户回复编号后,必须基于原始照片重新生成高清成品;不得裁切、放大或二次编辑宫格中的低清单格。
  6. 支持「再换一组」和「融合 02 和 05」;融合时先说明主风格与被吸收的特征,然后从原图生成。
  7. 用户明确说「跳过预览 / 直接出最终图」或已指定自选模板时,直达成品。只说「纸币」而未点名时,按当次需求选样张或实景。纸币规范见 banknote.md;中央邮票见 central-stamp.md;黑胶唱片见 vinyl-record.md。
  8. 用户说「先别生成」时,不生成图;最多用文字给 3 个方向。
  9. text 路径默认 preview_mode: skip。只有用户说「先给我几个方向」时,才出最多 4 个差异明显的隐喻方案(可文字,或 2×2 宫格)。
  10. 节日限定:用对话当天日期,窗口 ±1 天。默认 holiday_mode: auto(窗口内启用);用户说「不用节日限定」则 skip。不得为过节而给单人照补出伴侣。

宫格规则见 references/style-preview.md。
文本转译见 references/text-visual.md。
节日层见 references/holidays.md。

Decision priority

  1. 用户最新明确指令
  2. 明确视觉用途(封面 / Story / 海报等)
  3. 用户指定的 Style / Layout / Display
  4. 原图客观结构
  5. Skill 默认规则

Named template routing

用户说法style_mode核心关系
纸币样张 / 纪念钞样张banknote_specimen上方摄影,下方平放纪念券
纸币实景 / 手持纪念钞banknote_in_situ上方摄影,下方同场景手持纪念券
中央邮票 / 原位邮票central_in_place_stamp同一底图中央区域只改变媒介,不改变内容坐标
黑胶唱片 / 唱片店实景vinyl_record_in_situ上方正方形摄影,下方唱片店中的专辑实景
抽象色块 / 上下色块转译memory_color_blocks上方摄影,下方高概括色块
展览票exhibition_ticket摄影事实被编辑成收藏级票据

点名模板不是“风格预览候选”,而是明确创作指令。模板不得机械复用上一次生成的标题、编号、场景、媒介或构图。

Image analysis (internal)

无需逐项汇报,除非用户要求:

  • Subject:核心/次要主体、人物、建筑、自然、食物、标志物
  • Composition:视觉中心、视线/运动方向、景深、几何、可裁切区;判断主体 / 中景 / 远景是否可干净剥离
  • Emotion:宁静/松弛/孤独/烟火气等
  • Color:提取 3–6 主题色;可降饱和、合并;禁止无关彩虹堆色

Prompt construction order

  1. 基于输入图的二次创作
  2. 整体设计概念
  3. ORIGINAL_DISPLAY_MODE
  4. LAYOUT_MODE
  5. 原图裁切/保留方式
  6. 转译逻辑
  7. STYLE_MODE
  8. ABSTRACTION_LEVEL
  9. 色彩来源
  10. 留白与 translation scale
  11. 人物规则(若有)
  12. 文字
  13. 微图形(默认 1–3)
  14. 节日薄层(若窗口命中且未 skip)
  15. 背景材质
  16. 比例
  17. 禁止项

Strong defaults (override only with reason)

yaml
preset: default_editorial_memory
input_mode: auto                          # photo | text | mixed
preview_mode: auto                        # photo 未指定风格时先预览;text 默认 skip
preview_count: 6                          # 4 | 6 | 9
original_display_mode: split_top_bottom   # 或按图类型改,见 references
layout_mode: split_editorial
style_mode: auto                          # photo 先预览;text 默认 editorial_metaphor_card
abstraction_level: high                   # 保留约 15–30% 信息
whitespace_level: very_high
translation_scale: one_ninth_grid         # 大留白时转译约占一格 / 15–30% 面积
text_mode: auto_poetic                    # 8–20 字,不解释、不鸡汤
doodle_level: very_low
holiday_mode: auto                        # auto | force | skip;窗口 ±1 天
background_style: warm_off_white_paper
transition: deckled_paper_edge
ratio: 3:4

按图类型调整默认:自拍/人像 → taped_corner_photo,且转译禁止完整肖像;建筑 → 结构解构;极简几何 → extreme_minimal_abstraction;用户强调全新作品 → translation_only。完整规则见 references/defaults-and-presets.md。

Human / original / safety (hard rules)

  • 人物:保留姿态动作发型服装结构;简化五官与纹理;不得擅自改动作、加人、加道具、儿童化。
  • 禁止完整肖像画:人像转译不得变成可单独当肖像作品的写实水彩/线稿;五官最多 1–3 个记号。例外:banknote_specimen / banknote_in_situ 允许凹版雕线肖像,仍禁止照片级脸。
  • 原图出现在画面中时:可裁切缩放重组;不得改天气/人物/建筑/地貌(除非用户要求)。
  • 分层贴纸必须是 2–3 张可分离的白边贴纸,不得融成一张异形大贴纸。
  • 节日限定不得为过节而加人、改成情侣或改节日装;单人照只用低强度符号。
  • 不虚构未提供的地点、日期、票号、经纬度、真实机构标识;概念编号可用 NO. 001。
    节日层不代表照片拍摄于该日,禁止把「今天」写进伪造日期章。

Parameter cheat sheet

维度常用值详情
Inputphoto, text, mixedtext-visual.md
Displaysplit_top_bottom, taped_corner_photo, translation_only, …display-and-layout.md
Layoutsplit_editorial, large_whitespace_small_art, …同上
Stylebanknote_specimen, central_in_place_stamp, vinyl_record_in_situ, …styles.md
Previewauto / skip; 4 / 6 / 9 格style-preview.md
Holidayauto / force / skip;±1 天holidays.md
Abstractionlow / medium / high / extremesystems.md
Textnone, user_text, auto_poetic, …同上
Full schemaYAMLparameters.md

智能预设:travel_journal, personal_memory, one_day_exhibition, postcard_memory, through_glass, pure_memory, layered_sticker_memory, rounded_monoline_memory, editorial_text_card, banknote_specimen, banknote_in_situ, central_in_place_stamp, vinyl_record_in_situ → defaults-and-presets.md。

调用示例 → examples.md。
文本转译 → text-visual.md。
节日限定 → holidays.md。
纪念纸币 → banknote.md。
中央原位邮票 → central-stamp.md。 黑胶唱片实景 → vinyl-record.md。 质量清单与失败修正 → quality.md。


Minimal invocation

text
/visual-memory-translator
启用影像转译编辑器
@VisualMemoryTranslator

无其他参数且未指定风格时,先走 6 格风格预览。YAML 调用示例:

yaml
skill: visual_memory_translator
original_display_mode: taped_corner_photo
style_mode: extreme_minimal_abstraction
abstraction_level: extreme
text_mode: none
ratio: 3:4

Final instruction

Do not treat the reference image as mere style-transfer fodder.
Decide what deserves to survive, what should disappear, and whether the original remains visible—then reconstruct through editorial composition and controlled abstraction.

不是把照片重新画一遍,而是决定这张照片最终应该留下什么。
不是把句子画成故事,而是决定这句话只需要一个隐喻。

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Visual Memory Translator AI skill do?

Reinterprets photos into editorial, artist-book memory pages, and can also translate a sentence or concept into a sparse editorial metaphor card. Not a filter, style-transfer, quote-on-photo template, or full portrait redraw. Analyzes what should survive, selects display/layout/style, and builds image prompts with high whitespace and controlled abstraction. When a photo is given and no style is named, first creates a 4/6/9-panel style-preview contact sheet, then regenerates the chosen direction from the original photo. Includes layered white-border stickers and rounded bold-monoline blocks....

Why use Visual Memory Translator on TypingMind?

Because you install it once and use it with any model. Visual Memory Translator is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Visual Memory Translator in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/TanShilongMario/visual-memory-translator-SKILL/tree/main/visual-memory-translator. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Visual Memory Translator?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Visual Memory Translator?

As many as you like. As long as a model supports skills, you can use Visual Memory Translator with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Visual Memory Translator AI skill free?

Yes. It is published on GitHub by TanShilongMario under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇