Generate/edit images with Nano Banana Pro (Gemini 3 Pro Image). Use for image create/modify requests incl. edits. Supports text-to-image + image-to-image; 1K/2K/4K; use --input-image.
Install with Tessl CLI
npx tessl i github:Dicklesworthstone/pi_agent_rust --skill nano-banana-pro90
Does it follow best practices?
If you maintain this skill, you can automatically optimize it using the tessl CLI to improve its score:
npx tessl skill review --optimize ./path/to/skillValidation for skill structure
Draft-to-final image generation workflow
Draft uses 1K resolution
0%
100%
Final uses 4K resolution
0%
100%
4K not used for drafts
50%
100%
Filename timestamp format
0%
100%
Descriptive filename suffix
37%
100%
Absolute path to script
0%
100%
No cd to skill directory
100%
100%
uv run used
0%
100%
Multiple iteration steps
100%
100%
Without context: $0.2104 · 1m 2s · 17 turns · 61 in / 3,675 out tokens
With context: $0.3738 · 1m 11s · 18 turns · 2,240 in / 3,988 out tokens
Precise image editing with template and path handling
--input-image flag used
0%
100%
Editing prompt uses 'Change ONLY' structure
16%
100%
Editing prompt preserves other elements
58%
100%
Preflight input file check
100%
100%
No reading image back
100%
100%
Absolute path to script
0%
100%
uv run used
0%
100%
Filename follows timestamp pattern
0%
100%
edit_notes.txt explains approach
70%
100%
Without context: $0.5221 · 1m 58s · 25 turns · 117 in / 6,270 out tokens
With context: $0.4700 · 1m 35s · 26 turns · 2,246 in / 4,810 out tokens
Resolution mapping and API key argument handling
Thumbnail uses 1K
0%
100%
Web banner uses 2K
0%
100%
Print uses 4K
0%
100%
Uppercase K in resolutions
0%
100%
--api-key argument used
100%
100%
Filename timestamp format
0%
100%
Descriptive filename suffixes
100%
100%
Absolute path to script
0%
100%
uv run used
0%
100%
README documents resolutions
100%
100%
Without context: $0.2708 · 1m 17s · 20 turns · 24 in / 4,253 out tokens
With context: $0.4046 · 1m 23s · 21 turns · 2,243 in / 4,283 out tokens
Table of Contents
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.