banana-people

name: banana-people description: >- People & portrait photography via Google Gemini (Nano Banana Ultimate). Triggers on: portrait, headshot, lifestyle photo, people, person, editorial photo, team photo, professional photo, face, human. allowed-tools: Bash(uv:), Bash(ls:), Read

People & Portrait Photography

Input: $ARGUMENTS (optional — subject description and scene)

If $ARGUMENTS is provided, use it as the portrait brief and skip to prompt building. If empty, ask about the subject and setting.

Requirements

• uv installed
• GEMINI_API_KEY environment variable set (get one at https://aistudio.google.com/apikey)

Workflow

1. Understand — Parse the user's request. Identify the subject (age, build, expression), setting, and intended use. If the subject or intent is unclear, ask ONE clarifying question.
2. Build prompt — Construct a detailed Gemini prompt using the Prompt Formula below. Always describe clothing, environment, and camera distance explicitly.
3. Configure — Choose aspect ratio, resolution, and lens based on the Domain Defaults table. Use defaults unless the user specifies otherwise.
4. Generate — Run the script.
5. Deliver — Report the saved file path. Do NOT read the image file back. Offer to iterate: adjust pose, lighting, framing, or prompt.

Prompt Formula

Build every people/portrait prompt using this structure:

[Person: age/build/expression/clothing] + [Natural Action/Pose] + [Real-World Environment] + [Portrait Framing + Rule of Thirds] + [Natural Light Photography, specific lens]

Rules

• Positive framing: Describe what IS in the image, never what is absent.
• Strong verb opener: Start with Capture, Photograph, Shoot, Frame, Compose.
• Clothing and environment: Always describe what the person is wearing and the specific setting they are in.
• Camera distance: Specify explicitly — close-up, medium shot, full-body, three-quarter length.
• Mood keywords: Use "candid" for natural, spontaneous moments. Use "editorial" for styled, intentional compositions.
• No celebrity likenesses: Never reference real public figures. Describe features generically (e.g., "a person in their 30s with short dark hair").

Domain Defaults

Generation

uv run "${CLAUDE_PLUGIN_ROOT}/scripts/banana.py" \
  --prompt "<detailed prompt following the formula>" \
  --output "<YYYY-MM-DD-HH-MM-SS-descriptive-name>.png" \
  --resolution 2K \
  --aspect-ratio <ratio> \
  --model gemini-3-pro-image-preview

Editing (with input images)

When the user provides an existing photo to modify (e.g., change background, adjust lighting, swap clothing):

uv run "${CLAUDE_PLUGIN_ROOT}/scripts/banana.py" \
  --prompt "<edit instruction: what to change AND what to keep>" \
  --output "<YYYY-MM-DD-HH-MM-SS-descriptive-name>.png" \
  --input-image "/path/to/source.png" \
  --resolution 2K \
  --aspect-ratio <ratio>

For edits, describe both the change and what must stay identical:

"Change the background to a warm golden-hour park scene. Keep the subject's face, expression, clothing, and pose exactly the same."

Up to 14 input images can be passed (repeat --input-image for each).

Reference Prompts

See references/prompts.md for ready-to-use example prompts covering headshots, lifestyle, editorial, group, and environmental portrait scenarios.