Menu
Generate or edit images via Gemini image-capable models through aivault (no Gemini API key in skill runtime).
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Complete guide for building Moldable apps. Use this skill when creating or modifying apps, creating or managing Scheduled automations, implementing workspace-aware storage, using typed native APIs for media, location, clipboard, notifications, displays, shortcuts, power/session state, local authentication, haptics, secure storage, USB, HID, serial, MIDI, or Bluetooth, declaring nativeHardware permissions, integrating with Moldable desktop messages, publishing artifacts, configuring workspaces, managing skills/MCPs, or troubleshooting app issues.
Author, review, and template beautiful responsive presentations and interactive web artifacts in the Moldable Slides and Artifacts apps. Use when the user wants to create or edit a slide deck, pitch, talk, interactive report, calculator/chart/table experience, durable runtime state, deck/page template, or publishable artifact in Moldable. Covers optional deck runtimes, staged builds, responsive page/deck authoring, persistence, and the Slides/Artifacts RPC APIs instead of standalone files.
Complete guide for using aivault as a zero-trust local vault and proxy for API secrets. Use this skill when initializing/configuring aivault, managing secrets and credentials, invoking capability-backed API calls, setting workspace/group isolation, adding custom providers, or troubleshooting daemon and policy issues.
Search, scrape, crawl, map, parse, and operate Firecrawl browser/agent workflows through aivault-backed Firecrawl API capabilities. No Firecrawl API key is read by the skill runtime.
Transcribe audio with Deepgram through aivault capability json (no provider API key in skill runtime).
Transcribe audio with ElevenLabs speech-to-text through aivault capability json (no provider API key in skill runtime).
| name | nano-banana-pro |
| description | Generate or edit images via Gemini image-capable models through aivault (no Gemini API key in skill runtime). |
| compatibility | Requires `aivault` CLI and Node.js (runs via `npx tsx`). |
| credentials | [{"name":"GEMINI_API_KEY","description":"Google Gemini API key for image generation and editing","required":true}] |
This skill uses aivault to keep your Gemini API key safe. It calls the Gemini REST :generateContent endpoint through the gemini/models capability and saves the returned image.
Install aivault if it is not already available:
curl -fsSL https://aivault.moldable.sh/install.sh | sh
Store your Gemini API key in aivault. The credential and all gemini/* capabilities are auto-provisioned from the built-in registry.
aivault secrets create --name GEMINI_API_KEY --value "YOUR_API_KEY" --scope global
npx -y tsx {baseDir}/scripts/generate.ts --prompt "a cat eating a nano-banana" --filename ./out.png
# Higher resolution
npx -y tsx {baseDir}/scripts/generate.ts --prompt "a cozy reading nook" --filename ./out.png --resolution 2K
# Edit with input image(s)
npx -y tsx {baseDir}/scripts/generate.ts --prompt "make it look like a watercolor painting" --filename ./out.png -i ./in.png
# Raw upstream JSON
npx -y tsx {baseDir}/scripts/generate.ts --prompt "a cat eating a nano-banana" --filename ./out.png --json
gemini-3-pro-image-preview (override with --model).generationConfig.responseModalities=["TEXT","IMAGE"] and generationConfig.imageConfig.imageSize.