Source app creates the asset
The user finishes an image, short video, article, script, or design inside your product.
Developers
Connect your AI app to app-ready publishing workflows.
If you build an AI image generator, video maker, writing tool, design app, or content agent, your users eventually ask: where can this output go? AI PubKit gives you a destination registry, content package shape, and partner intake path for that handoff.
Customer test
The current product is useful, but it needs this developer handoff.
From a customer perspective, the app support pages explain destinations well. The missing step is how a source app sends generated images, videos, captions, prompts, and campaign context into the publishing workflow.
AI PubKit prepares destinations
The content package is mapped into app-specific captions, titles, descriptions, tags, media notes, and review checks.
Each destination gets routed safely
Direct paths are used where official platform rules allow it. Assisted drafts are created where approval or manual review is needed.
1. Build the package
Include source app details, generated text, asset URLs, prompt, language, destination list, and desired output types.
2. Read destination status
Use the public registry to decide whether the user can expect direct publishing, assisted drafts, or a researched future path.
3. Open composer or request review
Today, use the partner request path with the content package. Next, the same schema can power a hosted composer URL and ingestion API.
4. Route by platform rules
Keep platform approval, OAuth, media requirements, review state, and manual confirmation visible before a post is sent.
Example
What an AI video tool would send.
This is the handoff object a video or image app should be able to create after generation. The public schema defines the field names and allowed asset structure.
{
"schemaVersion": "0.1",
"sourceApp": {
"name": "VeoClip AI",
"website": "https://example.com/veoclip-ai",
"contentType": "ai-video"
},
"sourceContent": {
"title": "New product launch video",
"prompt": "Create a 12-second vertical product teaser for a skincare launch.",
"caption": "A short launch teaser for our new serum.",
"language": "en",
"assets": [
{
"type": "video",
"url": "https://cdn.example.com/video/launch-teaser.mp4",
"format": "mp4",
"aspectRatio": "9:16",
"durationSeconds": 12,
"mediaNotes": "Vertical teaser with product close-up, brand-safe text area, and no licensed music."
},
{
"type": "image",
"url": "https://cdn.example.com/image/cover.jpg",
"format": "jpg",
"purpose": "cover",
"altText": "Serum bottle on a clean bathroom counter with soft morning light.",
"coverText": "New serum launch"
}
]
},
"destinations": [
"TikTok",
"YouTube",
"Instagram",
"LinkedIn",
"WordPress",
"Telegram Channel"
],
"requestedOutput": [
"platform captions",
"video titles",
"hashtags",
"cover text",
"alt text",
"thumbnail notes",
"review checklist"
]
}Current status
Be honest about what is live today.
This keeps customer expectations clear while still giving developers a concrete way to start.
Live now
Public app registry, destination answer pages, content package schema, GitHub partner intake, and free planning tools. There is no production ingestion API yet.
Partner review
AI apps can submit their source app details, API docs, content package sample, destination needs, and testing availability.
Next product step
Hosted composer URL, signed package submission, OAuth destination connections, draft creation, and direct posting where official paths allow it.
Direct answer
How does an AI image or video app connect to AI PubKit?
Create a content package with generated media URLs, captions, prompts, metadata, and requested destinations. Check destination support in apps.json. Submit the package and source app details through the partner request path today; use the same package shape for the hosted composer or API when that product layer is added.