v1.0.1

Nano Banana Prompting Skill

BenignClawScan verdict for this skill. Analyzed May 1, 2026, 5:51 AM.

Analysis

The skill appears purpose-aligned for Gemini image prompting, but users should notice that it uses a Gemini API key and delegates actual image generation to a local helper script.

GuidanceThis skill looks safe to install if you intend to generate images with Gemini. Before using it, make sure the referenced nano-banana-pro or Gemini generator script is trusted, set your Gemini API key carefully, and avoid sending sensitive reference images unless you are comfortable with the provider handling them.

Findings (4)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

Abnormal behavior control

Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.

Agentic Supply Chain Vulnerabilities

SeverityLowConfidenceMediumStatusNote

README.md

**Requires:** The `nano-banana-pro` skill (bundled with OpenClaw) or any Gemini image generation script.

The artifact is instruction-only and delegates actual image generation to another local skill or script, so the safety of that helper matters.

User impactThe separate generator script will determine how prompts, reference images, output paths, and the Gemini API key are handled.

RecommendationInstall or use only a trusted, reviewed Gemini image-generation helper such as the expected bundled skill.

Unexpected Code Execution

SeverityLowConfidenceHighStatusNote

SKILL.md

uv run {nano-banana-pro-dir}/scripts/generate_image.py --prompt '<YOUR_JSON_PROMPT>' --filename "<descriptive-name>.png" --resolution 2K

The skill instructs the agent to run a local Python image-generation helper. This is disclosed and central to the stated image-generation purpose.

User impactWhen asked to generate an image, the agent may run a local helper command and create an output image file.

RecommendationConfirm the helper script is trusted and keep output filenames/path choices simple and intentional.

Permission boundary

Checks whether tool use, credentials, dependencies, identity, account access, or inter-agent boundaries are broader than the stated purpose.

Identity and Privilege Abuse

SeverityLowConfidenceHighStatusNote

SKILL.md

"requires": { "bins": ["uv"], "env": ["GEMINI_API_KEY"] }, "primaryEnv": "GEMINI_API_KEY"

The skill requires a Gemini API key, which is expected for Gemini image generation but still grants access to the user's provider account or quota.

User impactUsing the skill may consume Gemini API quota or incur provider-side usage, depending on the user's account.

RecommendationUse a dedicated or least-privileged Gemini API key where possible, monitor usage, and avoid sharing the key in prompts or files.

Sensitive data protection

Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.

Insecure Inter-Agent Communication

SeverityLowConfidenceMediumStatusNote

SKILL.md

For image editing (user provides a reference image): ... -i "/path/to/reference.png"

For edits, the workflow passes a user-provided reference image path to the Gemini image-generation helper, implying the image may be processed by the external generation service.

User impactPrivate prompts or reference images used for generation may be sent to or processed by the Gemini provider workflow.

RecommendationAvoid using sensitive personal or confidential reference images unless you are comfortable with the Gemini provider's handling of them.