sogni-gen 1.2.3 → 1.2.5

package/README.md

@@ -12,7 +12,15 @@ An [OpenClaw](https://github.com/OpenClaw/OpenClaw) plugin for AI image + video
 
 ### Quick Install (OpenClaw) - Recommended
 
-This repo ships an `openclaw.plugin.json` manifest so OpenClaw can automatically download and set everything up:
+Point your OpenClaw to the [`llm.txt`](https://raw.githubusercontent.com/Sogni-AI/openclaw-sogni-gen/main/llm.txt) and everything is set up — just paste the URL into Telegram, WhatsApp, or iMessages and the bot handles image and video generation automatically.
+
+```
+https://raw.githubusercontent.com/Sogni-AI/openclaw-sogni-gen/main/llm.txt
+```
+
+### Plugin Install
+
+This repo also ships an `openclaw.plugin.json` manifest so OpenClaw can automatically download and set everything up:
 
 ```bash
 # One command to install from GitHub
@@ -22,8 +30,6 @@ openclaw plugins install git@github.com:Sogni-AI/openclaw-sogni-gen.git
 openclaw plugins install sogni-gen
 ```
 
-That's it! OpenClaw will handle the rest.
-
 ### Manual Installation
 
 ```bash
@@ -46,6 +52,7 @@ If OpenClaw loads this plugin, `sogni-gen` will read defaults from your OpenClaw
         "config": {
           "defaultImageModel": "z_image_turbo_bf16",
           "defaultEditModel": "qwen_image_edit_2511_fp8_lightning",
+          "defaultPhotoboothModel": "coreml-sogniXLturbo_alpha1_ad",
           "videoModels": {
             "t2v": "wan_v2.2-14b-fp8_t2v_lightx2v",
             "i2v": "wan_v2.2-14b-fp8_i2v_lightx2v",
@@ -117,6 +124,10 @@ node sogni-gen.mjs -m flux1-schnell-fp8 "a dragon eating tacos"
 # JPG output
 node sogni-gen.mjs --output-format jpg -o dragon.jpg "a dragon eating tacos"
 
+# Photobooth (face transfer)
+node sogni-gen.mjs --photobooth --ref face.jpg "80s fashion portrait"
+node sogni-gen.mjs --photobooth --ref face.jpg -n 4 "LinkedIn professional headshot"
+
 # Image edit with LoRA
 node sogni-gen.mjs -c subject.jpg --lora sogni_lora_v1 --lora-strength 0.7 \
   "add a neon cyberpunk glow"
@@ -155,6 +166,26 @@ node sogni-gen.mjs --video --estimate-video-cost --steps 20 \
   -m wan_v2.2-14b-fp8_t2v_lightx2v "ocean waves at sunset"
 ```
 
+## Photobooth (Face Transfer)
+
+Generate stylized portraits from a face photo using InstantID ControlNet:
+
+```bash
+# Basic photobooth
+node sogni-gen.mjs --photobooth --ref face.jpg "80s fashion portrait"
+
+# Multiple outputs
+node sogni-gen.mjs --photobooth --ref face.jpg -n 4 "LinkedIn professional headshot"
+
+# Custom ControlNet tuning
+node sogni-gen.mjs --photobooth --ref face.jpg --cn-strength 0.6 --cn-guidance-end 0.5 "oil painting"
+
+# Custom model
+node sogni-gen.mjs --photobooth --ref face.jpg -m coreml-dreamshaperXL_v21TurboDPMSDE "anime style"
+```
+
+Uses SDXL Turbo (`coreml-sogniXLturbo_alpha1_ad`) at 1024x1024 by default. The face image is passed via `--ref` and styled according to the prompt. Cannot be combined with `--video` or `-c/--context`.
+
 Multi-angle mode auto-builds the `<sks>` prompt and applies the `multiple_angles` LoRA.
 `--angles-360-video` generates i2v clips between consecutive angles (including last→first) and concatenates them with ffmpeg for a seamless loop.
 `--balance` / `--balances` does not require a prompt and exits after printing current `SPARK` and `SOGNI` balances.
@@ -215,7 +246,10 @@ Multi-angle mode auto-builds the `<sks>` prompt and applies the `multiple_angles
 --auto-resize-assets  Auto-resize video reference assets
 --no-auto-resize-assets  Disable auto-resize for video assets
 --estimate-video-cost Estimate video cost and exit (requires --steps)
---ref <path|url>      Reference image for i2v/s2v/animate
+--photobooth          Face transfer mode (InstantID + SDXL Turbo)
+--cn-strength <n>     ControlNet strength (default: 0.8)
+--cn-guidance-end <n> ControlNet guidance end point (default: 0.3)
+--ref <path|url>      Reference image for i2v/s2v/animate/photobooth
 --ref-end <path|url>  End frame for i2v interpolation
 --ref-audio <path>    Reference audio for s2v
 --ref-video <path>    Reference video for animate workflows
@@ -236,6 +270,7 @@ Multi-angle mode auto-builds the `<sks>` prompt and applies the `multiple_angles
 | `chroma-v.46-flash_fp8` | ~30s | Balanced |
 | `qwen_image_edit_2511_fp8` | ~30s | Image editing with context |
 | `qwen_image_edit_2511_fp8_lightning` | ~8s | Fast image editing |
+| `coreml-sogniXLturbo_alpha1_ad` | Fast | Photobooth face transfer (SDXL Turbo) |
 | `wan_v2.2-14b-fp8_t2v_lightx2v` | ~5min | Text-to-video |
 | `wan_v2.2-14b-fp8_i2v_lightx2v` | ~3-5min | Image-to-video |
 | `wan_v2.2-14b-fp8_s2v_lightx2v` | ~5min | Sound-to-video |

package/SKILL.md

@@ -110,7 +110,10 @@ node sogni-gen.mjs -q -o /tmp/cat.png "a cat wearing a hat"
 | `--auto-resize-assets` | Auto-resize video assets | true |
 | `--no-auto-resize-assets` | Disable auto-resize | - |
 | `--estimate-video-cost` | Estimate video cost and exit (requires --steps) | - |
-| `--ref <path>` | Reference image for video | required for video |
+| `--photobooth` | Face transfer mode (InstantID + SDXL Turbo) | - |
+| `--cn-strength <n>` | ControlNet strength (photobooth) | 0.8 |
+| `--cn-guidance-end <n>` | ControlNet guidance end point (photobooth) | 0.3 |
+| `--ref <path>` | Reference image for video or photobooth face | required for video/photobooth |
 | `--ref-end <path>` | End frame for i2v interpolation | - |
 | `--ref-audio <path>` | Reference audio for s2v | - |
 | `--ref-video <path>` | Reference video for animate workflows | - |
@@ -134,6 +137,7 @@ When installed as an OpenClaw plugin, `sogni-gen` will read defaults from:
         "config": {
           "defaultImageModel": "z_image_turbo_bf16",
           "defaultEditModel": "qwen_image_edit_2511_fp8_lightning",
+          "defaultPhotoboothModel": "coreml-sogniXLturbo_alpha1_ad",
           "videoModels": {
             "t2v": "wan_v2.2-14b-fp8_t2v_lightx2v",
             "i2v": "wan_v2.2-14b-fp8_i2v_lightx2v",
@@ -177,6 +181,7 @@ Seed strategies: `prompt-hash` (deterministic) or `random`.
 | `chroma-v.46-flash_fp8` | Medium | Balanced |
 | `qwen_image_edit_2511_fp8` | Medium | Image editing with context (up to 3) |
 | `qwen_image_edit_2511_fp8_lightning` | Fast | Quick image editing |
+| `coreml-sogniXLturbo_alpha1_ad` | Fast | Photobooth face transfer (SDXL Turbo) |
 
 ## Video Models
 
@@ -206,6 +211,32 @@ node sogni-gen.mjs --last-image "make it more vibrant"
 
 When context images are provided without `-m`, defaults to `qwen_image_edit_2511_fp8_lightning`.
 
+## Photobooth (Face Transfer)
+
+Generate stylized portraits from a face photo using InstantID ControlNet. When a user mentions "photobooth", wants a stylized portrait of themselves, or asks to transfer their face into a style, use `--photobooth` with `--ref` pointing to their face image.
+
+```bash
+# Basic photobooth
+node sogni-gen.mjs --photobooth --ref face.jpg "80s fashion portrait"
+
+# Multiple outputs
+node sogni-gen.mjs --photobooth --ref face.jpg -n 4 "LinkedIn professional headshot"
+
+# Custom ControlNet tuning
+node sogni-gen.mjs --photobooth --ref face.jpg --cn-strength 0.6 --cn-guidance-end 0.5 "oil painting"
+```
+
+Uses SDXL Turbo (`coreml-sogniXLturbo_alpha1_ad`) at 1024x1024 by default. The face image is passed via `--ref` and styled according to the prompt. Cannot be combined with `--video` or `-c/--context`.
+
+**Agent usage:**
+```bash
+# Photobooth: stylize a face photo
+node {{skillDir}}/sogni-gen.mjs -q --photobooth --ref /path/to/face.jpg -o /tmp/stylized.png "80s fashion portrait"
+
+# Multiple photobooth outputs
+node {{skillDir}}/sogni-gen.mjs -q --photobooth --ref /path/to/face.jpg -n 4 -o /tmp/stylized.png "LinkedIn professional headshot"
+```
+
 ## Multiple Angles (Turnaround)
 
 Generate specific camera angles from a single reference image using the Multiple Angles LoRA:
@@ -319,6 +350,11 @@ ls -la ~/.clawdbot/media/inbound/*.jpg | tail -3
 cp ~/.clawdbot/media/inbound/<latest>.jpg /tmp/to_restore.jpg
 ```
 
+## IMPORTANT KEYWORD RULE
+
+- If the user message includes the word "photobooth" (case-insensitive), always use `--photobooth` mode with `--ref` set to the user-provided face image.
+- Prioritize this rule over generic image-edit flows (`-c`) for that request.
+
 ## Agent Usage
 
 When user asks to generate/draw/create an image:
@@ -336,6 +372,9 @@ node {{skillDir}}/sogni-gen.mjs -q --video --ref /path/to/image.png -o /tmp/vide
 # Generate text-to-video
 node {{skillDir}}/sogni-gen.mjs -q --video -o /tmp/video.mp4 "ocean waves at sunset"
 
+# Photobooth: stylize a face photo
+node {{skillDir}}/sogni-gen.mjs -q --photobooth --ref /path/to/face.jpg -o /tmp/stylized.png "80s fashion portrait"
+
 # Check current SPARK/SOGNI balances (no prompt required)
 node {{skillDir}}/sogni-gen.mjs --json --balance
 

package/llm.txt

@@ -138,6 +138,7 @@ node {{skillDir}}/sogni-gen.mjs --json --balance
 
 ## Agent Behavior Guidelines
 
+0. If the user includes the keyword "photobooth" (case-insensitive), always use `--photobooth` with `--ref` to the user face image. Do not fall back to `-c` edit flow for that request.
 1. When the user asks to "draw", "generate", "create", or "make" an image: generate an image and send it.
 2. When they ask to "animate", "make a video", or "create a video": use --video mode.
 3. When they send a photo and ask to edit/change/modify it: use -c with their image.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "sogni-gen",
-  "version": "1.2.3",
+  "version": "1.2.5",
   "description": "Sogni AI image generation plugin for OpenClaw",
   "type": "module",
   "main": "sogni-gen.mjs",

package/sogni-gen.mjs

@@ -593,7 +593,10 @@ const options = {
   refAudio: null, // Reference audio for s2v
   refVideo: null, // Reference video for animate workflows
   contextImages: [], // Context images for image editing
-  looping: false // Create looping video (i2v only): generate A→B then B→A and concatenate
+  looping: false, // Create looping video (i2v only): generate A→B then B→A and concatenate
+  photobooth: false, // Photobooth mode (InstantID face transfer)
+  cnStrength: null, // ControlNet strength override
+  cnGuidanceEnd: null // ControlNet guidance end override
 };
 const cliSet = {
   output: false,
@@ -630,7 +633,10 @@ const cliSet = {
   refImageEnd: false,
   refAudio: false,
   refVideo: false,
-  context: false
+  context: false,
+  photobooth: false,
+  cnStrength: false,
+  cnGuidanceEnd: false
 };
 
 // Parse CLI args
@@ -762,6 +768,15 @@ for (let i = 0; i < args.length; i++) {
   } else if (arg === '-c' || arg === '--context') {
     options.contextImages.push(args[++i]);
     cliSet.context = true;
+  } else if (arg === '--photobooth') {
+    options.photobooth = true;
+    cliSet.photobooth = true;
+  } else if (arg === '--cn-strength') {
+    options.cnStrength = parseFloat(args[++i]);
+    cliSet.cnStrength = true;
+  } else if (arg === '--cn-guidance-end') {
+    options.cnGuidanceEnd = parseFloat(args[++i]);
+    cliSet.cnGuidanceEnd = true;
   } else if (arg === '--last-image') {
     // Use image from last render as reference/context
     if (existsSync(LAST_RENDER_PATH)) {
@@ -830,6 +845,12 @@ Image Options:
   -c, --context <path>  Context image for editing (can use multiple)
   --last-image          Use last generated image as context
 
+Photobooth (Face Transfer):
+  --photobooth            Face transfer mode (InstantID + SDXL Turbo)
+  --ref <path|url>        Face image (required with --photobooth)
+  --cn-strength <n>       ControlNet strength (default: 0.8)
+  --cn-guidance-end <n>   ControlNet guidance end point (default: 0.3)
+
 Video Options:
   --video, -v           Generate video instead of image
   --workflow <type>     Video workflow: t2v|i2v|s2v|animate-move|animate-replace
@@ -885,6 +906,8 @@ Examples:
   sogni-gen --video --last-image "gentle camera pan"
   sogni-gen -c photo.jpg "make the background a beach" -m qwen_image_edit_2511_fp8
   sogni-gen -c subject.jpg -c style.jpg "apply the style to the subject"
+  sogni-gen --photobooth --ref face.jpg "80s fashion portrait"
+  sogni-gen --photobooth --ref face.jpg -n 4 "LinkedIn professional headshot"
 `);
     process.exit(0);
   } else if (!arg.startsWith('-') && !options.prompt) {
@@ -1142,6 +1165,8 @@ if (options._lastImagePath) {
     } else if (!options.quiet) {
       console.error('Warning: --last-image ignored for text-to-video workflow.');
     }
+  } else if (options.photobooth) {
+    if (!options.refImage) options.refImage = options._lastImagePath;
   } else {
     options.contextImages.push(options._lastImagePath);
   }
@@ -1156,6 +1181,14 @@ if (options.video) {
   if (!cliSet.timeout && !timeoutFromConfig && options.timeout === 30000) {
     options.timeout = 300000; // 5 min for video
   }
+} else if (options.photobooth) {
+  // Photobooth uses SDXL Turbo + InstantID ControlNet
+  options.model = options.model || openclawConfig?.defaultPhotoboothModel || 'coreml-sogniXLturbo_alpha1_ad';
+  if (!cliSet.width) options.width = 1024;
+  if (!cliSet.height) options.height = 1024;
+  if (!cliSet.timeout && !timeoutFromConfig && options.timeout === 30000) {
+    options.timeout = 60000;
+  }
 } else if (options.contextImages.length > 0) {
   // Use qwen edit model when context images provided (unless model explicitly set)
   options.model = options.model || openclawConfig?.defaultEditModel || 'qwen_image_edit_2511_fp8_lightning';
@@ -1176,6 +1209,18 @@ if (!options.video && (options.refAudio || options.refVideo || options.videoWork
   });
 }
 
+if (options.photobooth) {
+  if (!options.refImage) {
+    fatalCliError('--photobooth requires --ref <face-image>.', { code: 'INVALID_ARGUMENT' });
+  }
+  if (options.video) {
+    fatalCliError('--photobooth cannot be combined with --video.', { code: 'INVALID_ARGUMENT' });
+  }
+  if (options.contextImages.length > 0) {
+    fatalCliError('--photobooth cannot be combined with -c/--context.', { code: 'INVALID_ARGUMENT' });
+  }
+}
+
 if (options.video) {
   if (options.videoWorkflow === 't2v') {
     if (options.refImage || options.refImageEnd || options.refAudio || options.refVideo) {
@@ -2416,6 +2461,53 @@ async function main() {
       }
       
       await client.createImageEditProject(editConfig);
+    } else if (options.photobooth) {
+      // Photobooth: face transfer with InstantID ControlNet
+      log(`Photobooth with ${options.model}...`);
+      if (options.seed !== null && options.seed !== undefined) log(`Using seed: ${options.seed}`);
+
+      const faceBuffer = await fetchMediaBuffer(options.refImage);
+      const modelDefaults = getModelDefaults(options.model, openclawConfig);
+      const steps = options.steps ?? modelDefaults?.steps ?? 7;
+      const guidance = options.guidance ?? modelDefaults?.guidance ?? 2;
+
+      const projectConfig = {
+        modelId: options.model,
+        positivePrompt: options.prompt,
+        negativePrompt: '',
+        stylePrompt: '',
+        numberOfMedia: options.count,
+        tokenType: options.tokenType || 'spark',
+        waitForCompletion: false,
+        sizePreset: 'custom',
+        width: options.width,
+        height: options.height,
+        steps,
+        guidance,
+        disableNSFWFilter: true,
+        sampler: options.sampler || 'dpmpp_sde',
+        scheduler: options.scheduler || 'karras',
+        controlNet: {
+          name: 'instantid',
+          image: faceBuffer,
+          strength: options.cnStrength ?? 0.7,
+          mode: 'balanced',
+          guidanceStart: 0,
+          guidanceEnd: options.cnGuidanceEnd ?? 0.6,
+        }
+      };
+
+      if (options.outputFormat) projectConfig.outputFormat = options.outputFormat;
+      if (options.seed !== null && options.seed !== undefined) projectConfig.seed = options.seed;
+      if (options.loras.length > 0) projectConfig.loras = options.loras;
+      if (options.loraStrengths.length > 0) projectConfig.loraStrengths = options.loraStrengths;
+
+      const projectResult = await client.createImageProject(projectConfig);
+
+      // Check for errors in the response (e.g., insufficient tokens)
+      if (projectResult?.error || projectResult?.message) {
+        throw new Error(projectResult.error || projectResult.message);
+      }
     } else {
       // Standard image generation
       log(`Generating with ${options.model}...`);
@@ -2513,6 +2605,10 @@ async function main() {
       if (options.contextImages.length > 0) {
         renderInfo.contextImages = options.contextImages;
       }
+      if (options.photobooth) {
+        renderInfo.photobooth = true;
+        renderInfo.refImage = options.refImage;
+      }
       saveLastRender(renderInfo);
       
       // Save to file if requested
@@ -2700,6 +2796,15 @@ async function main() {
         if (options.contextImages.length > 0) {
           output.contextImages = options.contextImages;
         }
+        if (options.photobooth) {
+          output.photobooth = true;
+          output.refImage = options.refImage;
+          output.controlNet = {
+            name: 'instantid',
+            strength: options.cnStrength ?? 0.7,
+            guidanceEnd: options.cnGuidanceEnd ?? 0.6,
+          };
+        }
         console.log(JSON.stringify(output));
       } else {
         urls.forEach(url => console.log(url));