Meitu Visual Me

Always respond in the user's language.

Data Disclaimer:

• Local reads: Meitu Visual Me reads local files such as MEMORY.md, USER.md, and ./visual/PROFILE.md for personalized generation. This data never leaves your device.
• Uploaded to meitu API: Images you provide (photos, reference images) and generated prompts are sent to the Meitu OpenAPI for processing, subject to the Meitu privacy policy.
• First-use authorization: On first use, the skill will explain the above data usage. You can choose not to provide photos (falling back to text-to-image mode).

Overview

Memory-driven AI visual assistant. Reads user profiles and daily memories to generate personalized images and videos. Supports 17 scenario workflows and 28 trending styles. Play first, give feedback, and the system automatically learns your preferences.

You say	What it does
"帮我画" (draw something)	Automatically selects a scenario based on your current state and generates
Send a selfie + "换个赛博背景" (swap to cyber background)	Swaps the background, instant result
"头像系列" (avatar series)	Generates multiple style avatars from one face
"帮我做张 ID 卡" (make me an ID card)	Four trendy styles to choose from, exclusive collectible card
"微缩场景" (miniature scene)	Turns your current state into a toy miniature world
"一键换风格" (one-click style remix)	One image, multiple style variants
"今日卡" (daily card)	City + weather + mood, generates a daily info card
"帮我试穿这件" (try this on)	Send a clothing image for virtual try-on
"把这张图动起来" (bring this image to life)	Turns an image into a short video
"合个影" (take a photo together)	A photo with you and OpenClaw together

Dependencies

• tools: meitu-cli (npm install -g meitu-cli)
• credentials: meitu config set-ak --value <AK> + meitu config set-sk --value <SK>
• user knowledge (optional): ./visual/ (created on demand; if absent, all knowledge reads are skipped)
• project files: ./ (cwd, containing openclaw.yaml, DESIGN.md, etc.)
• platform context (optional): USER.md, MEMORY.md, memory/今日.md, SOUL.md, IDENTITY.md (platform-injected; skip if absent)
• user config (optional): $VISUAL/config/defaults.yaml

Path alias: In the text below, $VISUAL = {OPENCLAW_HOME}/workspace/visual/. When ./visual/ exists, it points to that directory; on the OpenClaw platform it auto-resolves to ~/.openclaw/workspace/visual/.

Platform Setup

OpenClaw: User data is located at ~/.openclaw/workspace/visual/. The platform injects it automatically; ./visual/ resolves to that path at runtime.

Claude Code / Other Agents: User data is located at ./visual/ under the current working directory, created automatically on first use.

All ./visual/ paths in the workflows below are resolved to the corresponding actual path by each platform at runtime.

Preflight

Must pass before every generation; if it fails, stop generation:

1. meitu --version — if not installed, prompt: npm install -g meitu-cli

2. Verify credentials: meitu auth verify --json — if credentials are invalid, this will return an auth error. For credential setup, read references/setup.md

3. Check whether the ./visual/ directory exists (if not, prompt first-time user; do not block)

4. Did the user send an image or a video? — Videos are not supported; prompt user to send a screenshot

5. One-time workspace resolution (results persist throughout the workflow): Detect mode: cwd has openclaw.yaml → project mode; else → one-off 检查 ./visual/ 目录 → 确定 capabilities

   Variables available after resolution (referenced directly in subsequent steps):

   mode             = "project" (cwd has openclaw.yaml) | "one-off"
   visual           = ./visual/ path (value if exists, otherwise null)
   can_read_knowledge = visual exists
   can_record       = mode == "project" AND visual exists
   output_dir       = Resolve output_dir: openclaw.yaml → "./output/" | else → "$VISUAL/output/meitu-visual-me/"
   

   mkdir -p {output_dir} Hard constraint: output_dir MUST NOT point to inside the skill folder

---

Core Workflow

Preflight → [Context] → Execute → [Refine] → Deliver → [Record]
              ↑ creative tasks run this       ↑ creative     ↑ project mode
              ↑ tool tasks skip               ↑ tool skip    ↑ one-off skip

Context

Part 1: Routing

Route first, read later — match the workflow first, then decide what to read and how to proceed.

First: Match Workflow

Compare the user's input against the trigger words in the Workflows section below to match a specific workflow. Read references/workflows.md for scenario details.

When no match is found, there are three cases:

• Clear description (e.g., "帮我画一只猫在月球上") — no workflow match needed; determine the style per prompt building rules, then generate directly with image-generate
• Only sent an image without text — recommend 4 suitable options based on image content (e.g., swap background, style remix, make ID card, image-to-video), let user choose
• Vague instruction (e.g., "帮我画", "帮我生张图") — recommend scenarios:
  1. Check ./visual/memory/global.md for preferred scenarios — if found, prioritize those
  2. No memory — pick 4 from the 17 workflows based on time of day and context, let user choose

Second: Determine Read Level and Style Handling

Workflow type	Read level	Style handling	Included workflows
Scenario generation	Standard read	Follow style rules (use memory if available, otherwise offer 4 options)	persona-diorama, daily-card, morning-grid, ootd, data-diorama, emotion-grid, relationship-board, memory-collage, real-toon
Guided generation	Standard read	Workflow includes built-in style selection step (e.g., ID card pick 1 of 4)	id-card, style-remix, avatar-series
Background swap	Light read	No style selection needed (scene described by user)	swap-bg
Tool operation	Zero read	No style involved, execute directly	beauty, virtual-tryon
Video	Light read	No style involved	to-video, motion-transfer

For image generation, read references/style-library.md for style keywords and auto-matching rules.

Third: Determine CLI Command

Each workflow's corresponding CLI command is listed in the Workflows section table. For detailed command parameters, read references/models.md. image-edit requires selecting a sub-model: praline (general editing) / gummy (portrait/pet photography, hairstyle adjustment).

Routing decision output (passed from Context to Execute):

{
  "workflow": "persona-diorama",
  "command": "image-generate",
  "read_level": "standard",
  "has_reference_image": true,
  "style": "微缩世界",
  "style_resolved": true
}

Part 2: Load Context

The read level is determined by the routing table above. Zero-read and light-read workflows do not need this step.

Standard read checklist (highest to lowest priority):

1. ./DESIGN.md — project context (if present, prioritize it; only supplement what it doesn't cover)
2. ./openclaw.yaml — resolve short-name references (e.g., brand: "acme" → ./visual/assets/brands/acme/)
3. ./visual/rules/quality.yaml — forbidden styles
4. ./visual/PROFILE.md — visual identity
5. ./visual/memory/global.md (first 30 lines) — preferences
6. Platform context (read if available, skip if not): USER.md, MEMORY.md, memory/今日.md, SOUL.md, Weather API

Use whatever you find; never fabricate user experiences.

Part 3: Build Prompt

Formula: [composition/camera angle] + [subject] + [action/expression] + [scene/background] + [style reference] + [lighting] + [user-specific details]

Hard rules:

1. Pure narrative style; keyword stacking is forbidden

   • ❌ woman, dress, red background, fashion, 8K, masterpiece
   • ✅ A young woman in a tailored brown dress, posing with confidence against a cherry red backdrop, fashion editorial feel with warm film grain

2. When a reference image is provided, explicitly state its purpose in the prompt (preserve facial likeness / reference composition / reference style); don't just attach an image without explanation

3. Embed style keywords into the narrative — pick 1-2 and weave them naturally into sentences; stacking at the end is forbidden

4. When user context is available, prioritize embedding personal details; when unavailable, generate directly based on user description without blocking.

5. Every generated image must have a clear style direction. Style determination priority:

   • User specified a style this time → use it directly
   • User didn't specify, but ./visual/memory/global.md has a recorded preferred style → use the memorized style directly without asking again
   • User didn't specify, and there's no memory → based on the day's state and scenario, pick 4 suitable styles from references/style-library.md's style matching guide, let user choose before generating

   Never generate a "generic" image without a style direction. If the user consistently picks the same style, remember it through the observation pipeline in the Record phase, and use it directly next time.

Example 1 (persona-diorama — from memory/今日.md, the user completed an important delivery today):

An isometric miniature workspace on a wooden desk. A tiny figure surrounded by [items extracted from memory]. A celebratory confetti cannon mid-burst on the desk corner. Warm afternoon light through a window, the whole scene feels like a lovingly crafted toy shop diorama.

Example 2 (swap-bg — user sent a selfie and said "换成东京街头"):

Keep the person exactly the same, change only the background to a vibrant Tokyo street scene with neon signs, busy crosswalks, and the warm glow of shop fronts lining both sides of the road, evening atmosphere with soft city lights

Example 3 (ID card — with reference image, user chose Cyber Neon style):

FACE LIKENESS IS THE #1 PRIORITY. Preserve exact facial structure. A young man with short black hair, black-framed glasses, and an oversized hoodie. Apply 3D collectible figure rendering LIGHTLY — keep proportions close to real. Generate a vertical ID CARD in Cyber Neon style: brushed carbon fiber base, neon cyan and magenta glow border, monospaced terminal font. Info area: "小明" / "独立开发者" / "杭州" — EST. 2000. Side strip: "卫衣收集者". Slight tilt, studio-lit product shot feel.

Execute

CLI command decision table:

Task type	Command	Key parameters	Notes
Text-to-image	meitu image-generate	--prompt --size --ratio	No reference image; --size accepts 2k/3k/WIDTHxHEIGHT
Generation with reference image	meitu image-generate	--image --prompt --size	Stylization / group photo
Background swap / content editing	meitu image-edit	--image --prompt --model	See model selection table
Face swap	meitu image-face-swap	--head_image_url --sence_image_url --prompt	Avatar series
Beauty enhance	meitu image-beauty-enhance	--image --beatify_type	Single-person photo
Image-to-video	meitu image-to-video	--image --prompt --video_duration	Async task
Motion transfer	meitu video-motion-transfer	--image_url --video_url --prompt	Async task
Virtual try-on	meitu image-try-on	--clothes_image_url --person_image_url --replace

image-edit model selection:

Priority	Model	Use case	Output style
1	nougat	Stylization: cartoon, 3D figure, anime, sketch, artistic recreation	Artistic (NOT realistic face)
2	gummy	Portrait/pet photography, hairstyle adjustment	Realistic portrait
3	praline (default)	Everything else: text manipulation, background swap, color grading, add/remove elements, multi-image fusion	General editing

Quick rule: "变画风" (style change, output doesn't look like real person) → nougat; "拍写真/换发型" (portrait photo) → gummy; "改内容" (modify content) → praline

Reference image routing:

Intent mode	Reference image handling	Command
Stylization (photo → figurine/anime/etc.)	Pass via --image, prompt describes target style	image-edit --model nougat (cartoon/anime/3D) or image-generate --image (general style transfer)
Background swap	Pass via --image, prompt explicitly says "keep person unchanged"	image-edit --model praline
Group photo (multiple references)	Pass multiple via --image, prompt describes group photo scene	image-generate --image img1 img2
Analyze reference only (not passed to tool)	Extract composition/style info into prompt	image-generate (pure text-to-image)
Face swap	Pass separately via --head_image_url and --sence_image_url	image-face-swap

Output directory: Use output_dir resolved in Preflight. Ensure the directory exists (mkdir -p). Add --json --download-dir {output_dir} to all meitu commands.

Result field references:

• With --download-dir → use downloaded_files[0].saved_path for local path (media_urls also available but local path is more reliable)
• Without --download-dir → use media_urls[0] for result URL
• Error response → check code and hint fields (NOT error_code / user_hint)

Image input: Pass user-provided images via URL or local path. When a reference image is needed, check ./visual/assets/references/user.jpg.

Command examples:

# Text-to-image
meitu image-generate --prompt "..." --size 2k --ratio 1:1 --json --download-dir {output_dir}
# Image editing
meitu image-edit --image <url> --prompt "..." --model praline --json --download-dir {output_dir}
# Image-to-video (async)
meitu image-to-video --image <url> --prompt "..." --video_duration 5 --json

Error degradation (try each level in order):

Level	Action	Example
L1	Remove low-priority modifiers	Drop lighting/material descriptions, keep subject + scene + style
L2	Downgrade enum parameters	--size 3k → --size 2k (image-generate); --size 2K → --size 1K (image-poster-generate); --ratio 9:16 → --ratio 1:1
L3	Remove optional inputs	Drop reference image, switch to pure text-to-image
L4	Minimize to core elements	Keep only subject + style, remove everything else
L5	Stop and report error	Inform user of the specific error, suggest checking credentials or contacting support

Escalate one level after 2 consecutive failures. For other errors, see references/troubleshooting.md.

Refine (MUST for creative tasks, skip for tool tasks)

Tool tasks (beauty, virtual-tryon, image-upscale) → skip this step and go directly to Deliver.

Iterative refinement loop for creative tasks:

1. Present result — show the generated image + briefly explain design rationale (why this style/composition/color palette)
2. Wait for feedback — three possible directions:
   • User approves ("好" / "不错") → proceed to Deliver
   • User requests modifications ("背景换一下" / "颜色太亮") → step 3
   • User rejects entirely ("完全不对" / "重新来") → go back to Execute and regenerate
3. Adjust and regenerate — modify prompt or parameters based on feedback → re-run CLI command → back to step 1
4. Iteration cap — recommend at most 3 rounds. After 3 rounds, proactively suggest: adjust the requirement direction, split into sub-tasks, or try a different workflow

Note: Do not generate multiple images consecutively without presenting results. Wait for user feedback after each generation.

Deliver

Files are already in the correct directory (Execute uses Preflight's output_dir); only rename is needed:

File renaming: mv {downloaded} {output_dir}/{YYYY-MM-DD}_{description}.{ext}

Path display rule: When presenting file paths to the user, prefer ~/.openclaw/... format. Some chat platforms (e.g., WeChat Work MEDIA) only recognize ~/ prefix paths; absolute paths will prevent images/files from displaying. Internal processing can use absolute paths.

WeChat delivery: For the WeChat channel (channel=openclaw-weixin), use MEDIA:visual/output/meitu-visual-me/{filename}.jpg (relative path).

After delivery: After delivery, wait for user feedback before the next generation. If user gives style/preference feedback, proceed to Record. When the user says "适配各平台", read references/channel-presets.md.

DESIGN.md maintenance: If DESIGN.md Iteration Log > 5 entries → compact: keep the most recent 5, archive older entries to ./drafts/design-history.md.

Record

can_record = false → skip this entire section. In one-off mode, all feedback is only effective for the current conversation.

Zero-write default — write nothing after generation; only write when user gives active feedback:

User says	What to do
"好" / "喜欢这个风格"	Append observation to ./visual/memory/observations/observations.yaml (auto, no user confirmation needed). If that observation's projects >= 2, mention non-blockingly at end of reply: "By the way, you've preferred X across N projects. Want to save it as a universal preference?" User confirms → write to memory/global.md or memory/scenes/{scene}.md, delete the observation; user ignores → do nothing
"不要 XX 风格"	Has openclaw.yaml → ask: "Only skip XX for this project, or never use XX for all future projects?" Project-only → append to ./DESIGN.md Constraints; all future → append to ./visual/rules/quality.yaml (requires user confirmation). No openclaw.yaml → applies only to current task, no write
"我是男的" (identity facts)	Write to ./visual/PROFILE.md
Sends a new reference photo	Save as ./visual/assets/references/user.jpg
"换个背景重生"	Re-run current task, no write

If the user says nothing, write nothing and don't read observations.yaml (zero overhead).

One-off mode exception: When a user repeatedly expresses the same preference in one-off mode (e.g., says "不要渐变" multiple times), the Agent MAY proactively suggest:

"You've mentioned no gradients several times. Want to add it to the global forbidden list?" → User agrees → write to $VISUAL/rules/quality.yaml (requires confirmation) → User disagrees → effective only for the current conversation

Positive feedback recording path:

1. Read $VISUAL/memory/observations/observations.yaml (create if it doesn't exist)
2. Scan for semantically similar keys → merge or create new
3. Write back to file
4. If len(projects) >= 2 → propose promotion (non-blocking)

Promotion proposal template:

"You've preferred X across N projects. Want to save it? → Save to {scope_hint} scene [default] → Save to global preferences → Don't save"

If scope_hint is not null → default to scenes/{scope_hint}.md; if null → default to global.md. User confirms → write to target file + delete observation entry. User ignores → do nothing.

For write format details, read references/feedback-loop.md when handling feedback; for observation lifecycle and classification, read references/memory-protocol.md

---

First-Time User Guide

When the ./visual/ directory doesn't exist, this indicates a new user. Read references/first-time-guide.md to execute the onboarding flow.

---

Workflows (17)

For all workflow details, read references/workflows.md

Generation

Trigger words	Workflow	Command
微缩场景 (miniature scene)、个人场景 (personal scene)	persona-diorama	image-generate
今日卡 (daily card)、城市打卡 (city check-in)	daily-card	image-generate
早安 (good morning)、晨间四宫格 (morning grid)	morning-grid	image-generate
OOTD、今天穿什么 (what to wear today)	ootd	image-generate
数据场景 (data scene)、今日数据 (today's data)	data-diorama	image-generate
九宫格 (nine-grid)、情绪九宫格 (emotion grid)	emotion-grid	image-generate
关系板 (relationship board)、关系拼贴 (relationship collage)	relationship-board	image-generate
记忆拼贴 (memory collage)	memory-collage	image-generate
合影 (group photo)、合个影 (take a photo together)	real-toon	image-generate
ID 卡 (ID card)、收藏卡 (collectible card)	id-card	image-generate
一键换风格 (one-click style remix)、换风格 (change style)、风格万花筒 (style kaleidoscope)	style-remix	image-edit --model nougat (cartoon/anime/3D) / image-edit --model praline (general) / image-generate

Editing

Trigger words	Workflow	Command
换背景 (swap background)、改背景 (change background)	swap-bg	image-edit --model praline
头像系列 (avatar series)、换头像 (change avatar)	avatar-series	image-generate → image-face-swap
美颜 (beauty)、磨皮 (skin smoothing)	beauty	image-beauty-enhance

Video & Try-on

Trigger words	Workflow	Command
动起来 (bring to life)、图生视频 (image to video)	to-video	image-to-video
做这个动作 (do this motion)	motion-transfer	video-motion-transfer
试穿 (try on)、试衣 (try clothes)	virtual-tryon	image-try-on

---

References

File	When to read
references/workflows.md	When executing a specific scenario workflow
references/models.md	When detailed command parameters are needed
references/style-library.md	When generating images (user-specified style or auto-matching needed)
references/channel-presets.md	When adapting for multiple platforms
references/feedback-loop.md	When handling user feedback write formats
references/memory-protocol.md	When handling observation lifecycle and preference classification
references/troubleshooting.md	When encountering errors
references/setup.md	For first-time configuration
references/first-time-guide.md	When ./visual/ doesn't exist (new user onboarding)