Skills nano-banana-edit
📷

nano-banana-edit

Safe 🌐 Network access⚙️ External commands

Edit Images with Google Nano Banana 2 on RunComfy

Also available from: agentspace-so,runcomfy-com

Need to edit product photos, swap backgrounds, or batch process images? Nano Banana 2 Edit on RunComfy offers batch support for up to 20 images with strong subject identity preservation and spatial edit controls.

Supports: Claude Codex Code(CC)
⚠️ 65 Poor
1

Download the skill ZIP

2

Upload in Claude

Go to Settings → Capabilities → Skills → Upload skill

3

Toggle on and start using

Test it

Using "nano-banana-edit". User asks to swap background to a rainy neon cyberpunk street while keeping the subject identity

Expected outcome:

The model generates 1-4 images with the same subject preserved against a rainy neon cyberpunk background, with consistent pose and framing.

Using "nano-banana-edit". User asks to edit multiple product photos with a warm studio sweep background

Expected outcome:

Batch of 3-5 product images with consistent warm grey studio backgrounds and subject positioning, ready for e-commerce listing.

Security Audit

Safe
v1 • 5/28/2026

Static analysis flagged 78 potential issues, but evaluation reveals all are false positives. The skill is a thin wrapper for the RunComfy CLI that invokes `runcomfy run google/nano-banana-2/edit` to call a legitimate image editing API. The "external commands" detections are markdown bash command examples, not executable code. The "path traversal" flags reference placeholder parameter examples (`<absolute/path>`) in documentation. The "weak crypto" detection is frontmatter YAML. Network access is scoped to RunComfy API endpoints only. Security controls are properly documented including JSON string transmission (no shell injection) and secure token storage with mode 0600.

1
Files scanned
186
Lines analyzed
8
findings
1
Total audits

High Risk Issues (3)

Static Analysis Pattern Detected: Weak Cryptographic Algorithm
Scanner flagged lines 4 and 162 as containing weak crypto. Evaluation shows this is a false positive: line 4 is YAML frontmatter with model documentation, line 162 is exit code documentation. No cryptographic operations present in this skill.
Static Analysis Pattern Detected: Path Traversal
Scanner flagged lines 71, 82, 95 as path traversal. Evaluation shows this is a false positive: these lines contain `<absolute/value>` style placeholder examples in documentation explaining CLI parameter types. These are not actual path traversal attempts.
Static Analysis Pattern Detected: Hidden File Access
Scanner flagged line 181 as hidden file access. Evaluation shows this is documented security practice: token is stored at `~/.config/runcomfy/token.json` with mode 0600 (owner-only read/write). This is intentional secure credential handling, not a vulnerability.
Medium Risk Issues (1)
Static Analysis Pattern Detected: External Commands
Scanner flagged 57 locations as 'Ruby/shell backtick execution'. Evaluation shows this is a false positive: these are bash command examples in markdown documentation blocks showing CLI usage syntax. The skill is a documentation wrapper, not executable code.
Low Risk Issues (2)
Static Analysis Pattern Detected: Hardcoded URLs
Scanner flagged multiple URLs as hardcoded. Evaluation shows these are legitimate API endpoints for the RunComfy Model API service being wrapped by this skill.
Static Analysis Pattern Detected: System Reconnaissance
Scanner flagged line 38 as system reconnaissance. Evaluation shows this is simple user intent detection for routing requests to the correct skill.
Audited by: claude

Quality Score

38
Architecture
100
Maintainability
87
Content
32
Community
66
Security
91
Spec Compliance

What You Can Build

E-commerce Product Photography

Batch edit product images with consistent backgrounds. Edit up to 20 SKU images in one call, preserving product identity while swapping studio backgrounds.

Marketing Campaign Creative

Create A/B variants for ad creative with seed-locked multiple outputs. Generate background swaps and localized edits for influencer and brand content.

Brand Asset Relocalization

Adjust brand assets for different markets while preserving composition. Consistent text and palette swaps across localized content variants.

Try These Prompts

Simple Background Swap
Keep the subject identity unchanged. Convert the background into [description of desired background].
Localized Object Edit
Remove the [specific object] only. Keep [list of elements to preserve], the [environment elements], and the lighting unchanged.
Batch Background Consistency
For each input image: keep the [subject identity/pose/clothing] unchanged. Convert the background to [description]. Center the subject at the same fraction of frame as the input.
Multi-Image Variation Set
For each input: preserve the [identity/brand element]. Replace [specific element] with [new description]. Maintain consistent lighting and color palette across all outputs.

Best Practices

  • Lead with preservation instructions, end with the change. Always state what to keep first, then describe the edit.
  • Use concrete spatial language for localized edits: 'background only', 'left object', 'upper-right corner'.
  • Lock aspect_ratio and resolution for batch consistency. Use the same prompt grammar across batch edits.

Avoid

  • Avoid long compound instructions (change A and B and C) which cause model drift.
  • Do not use passive voice for edit instructions. Be imperative: 'Replace X with Y' not 'the background should be changed'.
  • Do not skip preservation goals. State explicitly what to keep unchanged or the model will subtly modify faces and brands.

Frequently Asked Questions

How many images can I edit in one batch request?
You can pass 1 to 20 input images per API call. The first image is treated as primary with the rest providing auxiliary cues.
How many output images can I generate per request?
You can request 1 to 4 output images per call using the number_of_images parameter.
How do I preserve subject identity during edits?
Lead your prompt with preservation instructions: 'Keep the subject identity, pose, and clothing unchanged.' Then state what to change.
When should I use Nano Banana Edit vs other models?
Use Nano Banana Edit for batch edits, background swaps, and spatial localization. Use GPT Image 2 edit for multilingual text. Use Flux Kontext for single ref with precise local edits.
What file formats are supported for output?
Output formats include PNG, JPEG, and WebP. Resolution options range from 0.5K to 4K.
How does the skill handle API authentication?
Authentication uses runcomfy login for interactive sessions or RUNCOMFY_TOKEN environment variable for CI. Token is stored securely with mode 0600 permissions.

Developer Details

File structure

📄 SKILL.md