The Agent Skills Directory

[COMMAND_EXECUTION]: The skill constructs and executes a shell command using uv run python to run a local script .claude/scripts/nano_banana_pro.py. This is the primary function of the skill.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input via the prompt argument and interpolates it directly into a bash command string.
Ingestion points: User-provided image description in {{arg:1}}.
Boundary markers: The input is wrapped in double quotes within the bash command template.
Capability inventory: Uses uv run python to execute scripts with file system and network access (as permitted by allowed-tools).
Sanitization: No explicit sanitization or escaping of the user-provided prompt is defined in the markdown instructions, relying on shell quoting which can be bypassed if the input contains unescaped double quotes.

generate-image