AI agents use nano_banana_edit to create or update resources in Apple Shortcuts — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Apple Shortcuts environment.
Image editing is a write operation that modifies data reversibly. While editing can involve external API calls (Google Gemini), the operation itself does not execute arbitrary code, delete data irreversibly, or move money. Severity is medium because the blast radius of accidental image modification affecting user or business content is notable but not critical—most edits can be undone or restored from backups.
From the tool's definition Tool name 'nano_banana_edit' and description 'Edit an existing image using Google Gemini' indicates modification of image data. The verb 'edit' is a reversible write operation.
Documented attack patterns abuse exactly the kind of access nano_banana_edit gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Apple Shortcuts, and nothing reaches the server without passing your rules. This is the rule we recommend for nano_banana_edit:
{
"version": "1",
"default": "deny",
"tools": {
"nano_banana_edit": {
"limits": [
{
"counter": "nano_banana_edit_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}nano_banana_edit stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Edit an existing image using Google Gemini. It is categorised as a Write tool in the Apple Shortcuts MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Apple Shortcuts MCP server in PolicyLayer and add a rule for nano_banana_edit: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Apple Shortcuts. Nothing to install.
nano_banana_edit is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the nano_banana_edit rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for nano_banana_edit. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
nano_banana_edit is provided by the Apple Shortcuts MCP server (@mindstone/mcp-server-apple-shortcuts). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Apple Shortcuts, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
423 Apple Shortcuts tools catalogued and risk-classified — across an index of 43,000+ MCP servers.