You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
"description": "Configuration for the alt-text-scan plugin used by the GitHub Accessibility Scanner. See https://github.com/github/accessibility-scanner-alt-text-plugin#configuration for details.",
"type": "object",
"additionalProperties": false,
"properties": {
"$schema": {
"type": "string",
"description": "Optional reference to this schema for editor-time validation. Ignored at scanner runtime."
},
"rules": {
"type": "object",
"description": "Per-rule enable/disable overrides. Rules not listed keep their default behavior (enabled with the exception of alt-text-quality).",
"additionalProperties": false,
"properties": {
"missing-alt-text": {
"type": "boolean",
"description": "Flags <img> elements where the alt attribute is absent or whitespace-only. alt=\"\" is treated as intentional decorative use and is not flagged."
},
"vague-alt-text": {
"type": "boolean",
"description": "Flags alt text that is a generic single word (image, photo, icon, logo, screenshot, etc.) or short filler phrase (an image of, a photo of)."
},
"filename-alt-text": {
"type": "boolean",
"description": "Flags alt text that ends in a common image file extension (.png, .jpg, .jpeg, .gif, .svg, .webp, .bmp, .ico)."
},
"placeholder-alt-text": {
"type": "boolean",
"description": "Flags alt text matching a known boilerplate string (todo, tbd, fixme, placeholder, alt text, insert alt text, image alt)."
},
"repeated-alt-text": {
"type": "boolean",
"description": "Flags two or more adjacent images in the rendered page that share the same normalized alt text."
},
"alt-text-quality": {
"type": "boolean",
"description": "Opt-in, model-backed. Sends each image and its alt text to a vision model (GitHub Models) to flag quality issues such as redundant prefixes, vague or context-missing descriptions, and link text that announces the destination's medium (e.g. 'PDF of…', 'Link to…') instead of naming it. Disabled by default; requires GITHUB_MODELS_TOKEN and incurs per-image API cost."