Chrome Extension v6.7.1

Scrya — Grok Chrome
Extension & Prompt Enhancer

The Chrome extension for AI prompt injection. Works with Grok, Google Gemini, Meta AI, ChatGPT, Midjourney, Leonardo, and more — 40+ prompt packs, 8,000+ prompts, Victoria character chat, batch generation, and a full creative pipeline.

Sign in to Chrome before visiting the Web Store. If you see "This item is not available", you need to sign in first.

Chrome Web Store sign-in required screenshot
16 Creative Tabs
40+ Prompt Packs
8,000+ Prompts
28 Languages
Beatles diorama — isometric 3D scene generated with Grok Imagine using the Scrya prompt pipeline

Generated with Grok Imagine using the Scrya prompt pipeline — isometric, 3D view, photorealistic

Core Capabilities

Everything you need to enhance prompts, extract media, and build datasets from Grok Imagine.

Media Extraction

Extract from Liked, Created, Saved, and Bookmarked sources with multi-pass farming and automatic deduplication.

  • Multi-pass pagination across 4 sources
  • Automatic deduplication by post ID
  • Rate-limited requests to avoid throttling

Cloud Sync

Upload media to Cloudflare R2 via server-side presigned URLs with Supabase as the metadata backend.

  • Server-side presigned URLs (no client secrets)
  • Per-user R2 key isolation
  • Background sync with progress toast

Character Builder

Sims-style interactive character creator with 25 configurable attributes across 7 sections and live prompt preview.

  • 10 skin tones, 15 hair colors, 8 eye colors
  • Save characters to cloud for reuse
  • Randomize with weighted distributions
Marketplace

Create & Browse Prompt Packs

Create your own packs or browse 40+ curated packs with 8,000+ prompts. Pick a type (Actions, Camera, Dialogue, Effects, Outfits, Lighting, Locations, Characters), type a theme, and AI generates a complete pack. Import from .txt files too.

  • 40+ packs with 8,000+ prompts across 83 categories (UNSPC taxonomy)
  • 3,000+ clothing prompts across 16 fashion categories
  • Cross-pack affinity: activate "Courtroom" actions, get matching locations, outfits, lighting
  • Publish for the community or keep private
Scrya — Create Prompt Pack wizard with 8 prompt types for Video and Image

5-step pack creation wizard — built right into the extension

Scrya extension in Japanese — Karate-Goju pack with localized sidebar

Localized in 28 languages — Japanese locale shown with Karate-Goju pack

Built for Shopify

Shopify Integration

Connect your Shopify store and generate AI product images using prompt packs. Push generated images directly to your product listings.

1
Connect Your Store
Paste your Shopify access token — no OAuth app needed. Tokens stay server-side.
2
Import Products
Browse your product catalog inside Scrya. Images, variants, and tags synced.
3
Generate AI Images
Pick a prompt pack, generate styled product images with Grok or ComfyUI.
4
Push to Shopify
One click pushes generated images to your Shopify product listing.
Scrya outfits tab — 3,000+ clothing prompts across 16 categories with AI thumbnails

3,000+ outfit prompts with AI-generated thumbnails across 16 fashion categories

Challenge Entry
Live
J
Jane D.
2 min ago
"A samurai standing at the edge of a misty cliff..."
42 likes Share
scrya.com/entry/abc123 · auto-copied on submit
Community

Challenge System

Submit challenge entries directly from prompt cells, share them with OG-tagged links, and discover what other creators are making. Each entry gets a public page with video player and likes.

  • Submit entries from any prompt cell in the grid
  • Shareable links at scrya.com/entry/{id} with OG tags
  • Creator profiles with avatar, bio, and gallery
  • Entry ticker showing recent submissions in real time
Character Chat

Victoria — AI Character Chat

Chat with Victoria, an AI character who lives inside the extension. She remembers your conversation, suggests prompts mid-chat, speaks in your language, and can end a scene by injecting a generated video prompt directly into Grok Imagine.

  • 9 supported languages — en, es, fr, de, pt, ja, ko, zh, ar
  • Multi-beat burst bubbles — replies pop in one sentence at a time with a typing indicator
  • End Turn — she can autonomously inject a scene prompt or end the conversation
  • Import any IMDb character as a chatbot persona — search, seed, and chat in seconds
  • Eleven Labs TTS — hear replies with 150+ voices; SSML prosody control
V
Victoria
● Online
What should we shoot next?
V
Something cinematic — misty forest at dawn.
V
Slow pan. Dew on the leaves.
V
Reply to Victoria…
Resilient Injection

Target Picker

A visual element picker like Chrome DevTools, purpose-built for prompt injection. Hover to highlight text inputs on grok.com, click to save the CSS selector. When Grok changes their UI, just re-pick the target.

  • Visual hover-to-highlight element picker
  • Crosshair button in the pipeline toolbar
  • Survives Grok UI changes -- just re-pick
  • Auto-retry with fallback to auto-detect
Target Picker
Pick Target Element
textarea.prompt-input
Hover over elements on grok.com to highlight. Click to save selector.
Saved selector active div[data-testid="prompt"]
Scrya command palette — search and queue prompts for batch generation

VS Code-style command palette with batch queue — watch the "Newsletter Blast" prompt in action

Automation

Command Palette + Batch Queue

Hit / to open the command palette and search across all prompt packs instantly. Queue up prompts and batch-generate them — Scrya drives grok.com's UI sequentially with zero API costs.

  • VS Code-style search across all packs with thumbnails
  • Configure mode (image/video), aspect ratio per batch
  • Auto-download completed generations
  • No API key needed — uses free Grok web interface
Filter Rules
3 active
Landscapes
mountain, valley, sunset, horizon
47 matched
Portraits
face, portrait, headshot, close-up
123 matched
Cyberpunk
neon, cyber, futuristic, hologram
31 matched
Organization

Smart Categorization

Define keyword-based filter rules that automatically organize your extracted media into lists. Run rules individually or in batch to keep your dataset structured.

  • Run individual rules with detailed match stats
  • Dual-write persistence to Chrome storage and Supabase
  • Automatic migration from local to cloud on first auth
Creative Toolkit

16 Specialized Tabs

A dedicated tab for every dimension of your creative pipeline, from action prompts to lighting presets to simulation worlds.

01
Favorites
Extracted media
02
Actions
Video action prompts
03
Camera
Angles and movements
04
Storyboard
Direction and NCP
05
Effects
377 visual effects
06
Outfits
Wardrobe prompts
07
Lighting
Light and mood
08
Locations
290 scene locations
09
Characters
25-attr builder
10
Dialogue
Speech and text
11
Prompts
Custom prompt editor
12
Wildcards
Random modifiers
13
Generator
Batch generation
14
Feed
Challenge entries
15
SimWorld
Simulation viewer
16
Settings
Config and sync
Pipeline Chips

Interactive chiplets embedded in the toolbar that change based on the active tab. Build your prompt visually by toggling chips on and off.

cinematic lighting portrait silk dress golden hour slow motion mountain lake cyberpunk film grain
Studio Alpha
3 members
Team
AC
Alice Chen
Owner
Owner
BM
Bob Martinez
Editor
Editor
Viewer
Collaboration

Team Workspaces

Create teams, invite members, and share datasets. All data is workspace-scoped so team members see the same collections, favorites, and filter rules.

  • Role-based access: owner, admin, editor, viewer
  • Invite by email or User ID
  • Workspace selector in Settings to switch contexts
  • R2 keys scoped under team_{id}/ for isolation
Multi-Platform

10 Supported AI Platforms

Prompt injection works across all of these — same one-click workflow, different tab.

Grok / Grok Imagine Primary
Full batch queue + media extraction
Google Gemini
rich-textarea Angular injection
Meta AI
Lexical contenteditable injection
ChatGPT
ProseMirror injection
Midjourney
Discord-style input
Leonardo.ai
Generation page detection
Google ImageFX
ImageFX prompt field
Automatic1111 Local
txt2img + img2img tabs
ComfyUI Local
Node workflow injection
Artlist
AI video/image generation

+ unlimited custom sites via the Target Picker — hover to highlight, click to save the CSS selector

Changelog

Release Notes

Every version ships real features. No marketing bumps.

v6.7.1 Current
2026-06-25
  • Multi-beat replies — longer character replies split into burst bubbles with "typing…" indicator between each one
  • Sound + Jailbreak consolidated into the brain button settings tab
  • End Turn category filter — conflict categories no longer bleed into romance-world action picker
  • Synchronized TTS audio: sound plays on each beat as it lands, not at reply end
v6.6.0
2026-06-15
  • Eleven Labs TTS — 150+ community voices in a searchable, filterable browser; SSML prosody support
  • IMDB film import — search TMDB by title, seed story DAG via Temporal workflow, import characters as chatbots; background polling survives dialog close
  • End Turn actions — Victoria can autonomously act, inject a video prompt, or end the scene at end of reply
  • Voice-clone chat replies using your Eleven Labs voice library
v6.5.0
2026-05-20
  • Victoria character chat with roleplay in 9 languages (en, es, fr, de, pt, ja, ko, zh, ar)
  • Method-acting persona brief — LLM-expanded character system prompts from IMDb story DAG fields
  • Scrya LLM gateway preset — one-click configure against api.scrya.com
  • Choice buttons restyled; bubble shows only the dialogue text (not full reply)
v6.0.64
2026-04-17
  • Prompt injection into Google Gemini (gemini.google.com) — same workflow as Grok and Meta AI
  • content_gemini.js: targets rich-textarea Angular contenteditable; execCommand injection for reliable framework detection
  • Gemini added to SPA navigation detection, webRequest monitoring, INJECTION_SITES config
v6.0.63
2026-04-15
  • Victoria chat welcome screen surfaces A1111/ComfyUI CORS setup tips
  • Skip context dialog on web — silent CORS probe replaces blocking popup
  • Fixed resume greeting race condition with isInitialized flag
  • Claude Code API: direct HTTP on web, bypassing extension background
  • Removed Paint that scene button from chat toolbar
v6.0.62
2026-04-10
  • Batch generation queue extended to Meta AI and ChatGPT
  • Batch button picker lets you choose target site per queue run
  • Progress detection: SVG circle dashoffset + percentage text + spinner element fallbacks
v6.0.26
2026-03-12
  • Batch Generation Queue — client-side DOM automation for sequential prompt generation (no API costs)
  • Grok Chat Tab — chat with Grok via xAI API with auto-screenshot context
  • Queue Processor (content.js): SVG stroke-dashoffset progress polling
  • Auto-download on batch completion

Start Enhancing Your Prompts

Install the extension, sign in with Google, Twitter, or GitHub, and inject into Grok, Gemini, ChatGPT, and more in seconds.