default_enabled = false
manifest_label = "Image generation"
manifest_description = "Create images with a supported OpenAI or OpenRouter model"
render_pending = "Generating image"
render_ready = "Image ready"
render_failed = "Image generation failed"
model = "gpt-image-2.5-sunburst"
[tool]
name = "generate_image"
description = "Generate one image from a prompt, optionally using images already attached to this chat as references. The finished image is sent to the user."
[tool.parameters]
type = "object"
required = ["prompt"]
additionalProperties = false
[tool.parameters.properties.prompt]
type = "string"
minLength = 1
maxLength = 32000
description = "Describe the image to create or how to edit the reference images."
[tool.parameters.properties.image_aspect]
type = "string"
enum = ["square", "landscape", "portrait"]
description = "Choose the canvas shape for the image. Omit for square (1:1)."
[tool.parameters.properties.reference_file_ids]
type = "array"
maxItems = 4
items = { type = "string" }
description = "Optional authorized image file IDs from this chat."