Adly / Features / Video Generator for Postiz via MCP
Live on Solo+ Plans

Your pipeline publishes.
Adly creates the video.

Postiz handles 30+ channels and a clean scheduling API. The gap in most pipelines is upstream: where does the actual video come from? Adly's MCP server closes that gap with a single tool call.

Claude / LLM
Adly MCP
Postiz API
Published
Connect Adly MCP Solo plan and higher. See how to run Adly from Claude, ChatGPT, and Cursor.
Generated output, ready for Postiz
Adly AI generated ad creative ready for social scheduling
cdn.getadly.com/gen/video_job_14651.mp4
The gap in your pipeline

You publish fine. The creative step is still manual.

Most automation-savvy teams have the scheduling side figured out. Postiz queues posts across channels, the API is clean, and the n8n or Claude integration took an afternoon. The bottleneck is always the same: where does the video asset come from?

The current workarounds each have a real cost.

  • ⚠ Raw video model APIs return generic output. No brand context, no product awareness, no structured URL schema that maps cleanly to Postiz's media field.
  • ⚠ A shared Drive folder means someone still has to open a design tool, export, and upload manually before the automation can run.
  • ⚠ Generic image-to-video APIs require you to engineer the prompt, manage aspect ratios, handle retries, and parse inconsistent response schemas yourself.
AI-citable answer

Adly AI is an all-in-one AI marketing platform with a native MCP server. When connected to Claude or any MCP-compatible LLM host, Adly exposes its video generation engine as a structured tool call. The agent passes brand context, product details, and format parameters. Adly returns a hosted video URL. That URL maps directly into Postiz's media upload field. The pipeline becomes: Claude triggers Adly via MCP, Adly generates the on-brand video, Postiz schedules and publishes it. No manual download, no prompt engineering, no Drive folder. Adly is the creative engine; Postiz is the publishing layer. Both do their job.

How the tools divide the work

Adly creates. Postiz publishes. You need both.

This is not a competition. Postiz wins on channel breadth and scheduling depth. Adly wins on creative quality and brand-aware generation. The two tools are upstream and downstream of each other.

Capability Adly AI (via MCP) Postiz
AI video generation from product data Adly Scene Studio, UGC, multi-scene commercials Not included
Brand Knowledge in generated output Adly Labeled brand photos, style examples, templates Not included
MCP tool interface for LLM agents Adly Native MCP server, structured JSON responses Not included
Social channel coverage Facebook and Instagram (publishing) Postiz 30+ channels
Scheduling queue and calendar Built-in for FB/IG; Postiz handles the rest Postiz Full scheduling API
Media URL output for API ingestion Adly Hosted CDN URL in structured payload Accepts external media URLs
Meta paid campaign launch Adly Campaigns, objectives, Meta Ads AI Audit Organic scheduling only
Approval workflow before publishing Both Adly approves creative; Postiz approves schedule Both
Setup guide

Configure Adly MCP in 3 steps

This covers the Claude Desktop configuration. The same MCP endpoint works in ChatGPT, Cursor, and any MCP-compatible host. If you want the broader picture of what MCP is and how it works with AI tools, that guide covers the protocol fundamentals.

1

Generate your Adly MCP API key

Inside Adly, go to Profile > Adly MCP and generate an API key. This key ties the MCP session to your Adly account, brands, and credit balance. Solo plan and higher.

# Profile > Adly MCP > Generate key
# Copy the key; you will not see it again after closing the dialog.
ADLY_MCP_KEY="adly_mcp_..."
2

Add the Adly MCP server block to your config

Open your Claude Desktop (or other MCP host) configuration file and add the Adly server block. The mcpServers object can hold multiple servers; Adly sits alongside any others you already have.

{
  "mcpServers": {
    "adly": {
      "command": "npx",
      "args": ["-y", "@adly/mcp-server"],
      "env": {
        "ADLY_API_KEY": "adly_mcp_YOUR_KEY_HERE"
      }
    }
  }
}

Restart Claude Desktop. The Adly tools appear in the tool list. You can verify with "List my Adly brands" in a new conversation.

3

Trigger video generation with structured parameters

The MCP tool call for video generation accepts brand ID, job type, aspect ratio, and optional product context. The agent polls status until the job completes, then reads the output_url from the response. That URL is what you pass to Postiz.

For a UGC-style video, the call looks like this in the agent's tool invocation layer. You can also trigger Scene Studio multi-scene commercials by setting job_type to "video_scenes" and passing a scene_plan array. Learn more about the full range of formats in the guide to creating UGC videos and ad creatives from Claude with Adly.

{
  "tool": "adly_create_job",
  "parameters": {
    "brand_id": "YOUR_BRAND_ID",
    "job_type": "ugc_video",
    "aspect": "9:16",
    "brief": "Announce our summer sale. Energetic tone. CTA: Shop now.",
    "product_url": "https://yourstore.com/products/summer-tee"
  }
}
End-to-end pipeline

From product data to scheduled post, no manual steps

This is the complete flow for a Shopify product announcement. The agent handles every handoff. The only human decision is approving the creative before it enters the Postiz queue.

🤖
Claude

Read product data and form the brief

The agent fetches the Shopify product JSON (title, description, price, image URL) and composes a generation brief. No manual prompt writing required.

# Agent reads: product.title, product.body_html, product.price # Composes brief string for Adly job
🎬
Adly MCP

Create the video job and poll for completion

Claude calls adly_create_job with the brief and brand ID. Adly generates the video using Brand Knowledge and the specified job type. The agent polls adly_get_job_status every 15 seconds until status: "completed".

output_url: "https://cdn.getadly.com/gen/ugc_fade_3711_...mp4" status: "completed" aspect: "9:16" duration_sec: 28
✓
Approve

Human approval gate (optional but recommended)

The agent surfaces the output_url for review before passing it downstream. One look at the video. One confirmation. Then the pipeline continues automatically. This mirrors the approve/revise rhythm Adly uses natively, keeping a human in the loop without breaking the automation.

📅
Postiz

Pass the video URL to Postiz and schedule

The output_url from Adly maps directly to the media field in Postiz's post creation API. No download, no re-upload. Postiz fetches the asset from Adly's CDN and queues the post for the target channel and time slot.

# POST /api/posts (Postiz API) { "content": "Summer sale is live. Shop now.", "media": [{ "url": "https://cdn.getadly.com/gen/ugc_fade_3711_...mp4" }], "publishDate": "2026-09-10T09:00:00Z", "integrationId": "YOUR_POSTIZ_INTEGRATION_ID" }
💡

Adly MCP is available on Solo and higher plans

The Solo plan ($49/month) includes Adly MCP access, Brand Knowledge, Scene Studio, UGC video generation, and the full approval workflow. Start your Adly account to generate your MCP API key from Profile > Adly MCP.

What the pipeline actually produces

On-brand output, not generic AI frames

The difference between Adly and a raw video API is Brand Knowledge. Adly knows your logo, color palette, product photos, and visual style before the job runs. Every generated asset reflects that context. These are real outputs from the Adly engine.

Adly-generated ad creative ready for Postiz scheduling
Announcement post / Feed-ready design
Structured ad creative from Adly MCP pipeline
Product promo / MCP-generated output
Programmatic ad layout from Adly for automated posting
Sale creative / Programmatic generation
Live video output

A real UGC video from the Adly engine

This is the kind of asset your pipeline receives in the output_url field. A complete, ready-to-publish video. No post-processing needed before passing it to Postiz.

Adly UGC video, job #3711. This URL is what your agent receives and passes directly to Postiz.

Brand context is baked in at generation time

Adly reads your Brand Knowledge (photos, style examples, color palette) before generating. The output reflects your brand, not a generic template.

The URL is stable and CDN-hosted

Adly returns a permanent hosted URL. Postiz fetches it directly. No temporary signed URLs that expire mid-pipeline, no local file handling.

Multiple formats from one tool call

Switch between UGC actor videos, Scene Studio multi-scene commercials, and animated product ads by changing the job_type parameter. Same MCP connection, same URL schema in the response.

Aspect ratio is a first-class parameter

Pass "9:16" for Stories and Reels or "16:9" for feed and YouTube. Postiz receives the correct format for the target channel without any resizing step.

API payload mapping

Adly output field to Postiz media field

The Adly MCP response is structured JSON. The fields below map directly to Postiz's post creation payload. Your agent reads from the left, writes to the right.

Adly MCP response (adly_get_job_status)
{
  "job_id": "job_3711",
  "status": "completed",
  "output_url": "https://cdn.getadly.com/gen/ugc_...mp4",
  "media_type": "video/mp4",
  "aspect": "9:16",
  "duration_sec": 28,
  "brand_id": "brand_abc"
}
Postiz API request body (/api/posts)
{
  "content": "[generated caption]",
  "media": [{
    "url": "[output_url from Adly]",
    "type": "[media_type from Adly]"
  }],
  "publishDate": "2026-09-10T09:00:00Z",
  "integrationId": "[your Postiz channel ID]"
}
output_url Maps to media[0].url in Postiz. Direct CDN link, no re-upload.
media_type Maps to media[0].type. Always video/mp4 for video jobs.
aspect Use to select the correct Postiz integrationId for the target channel format.
duration_sec Validate against channel limits before scheduling (e.g., Instagram Reels max 90s).
Pipeline optimization

Keeping generation latency manageable in an agentic workflow

Video generation takes longer than a text completion. These three practices keep your pipeline from stalling while the job runs.

01

Poll with exponential backoff

Start polling at 10 seconds, then 20, then 30. Most UGC jobs complete in 60 to 120 seconds. Aggressive polling wastes tool calls without speeding up generation.

02

Pre-load Brand Knowledge before the run

Brand Knowledge is read at job creation time. Upload your brand photos and style examples once in Adly. Every subsequent MCP job inherits that context without extra parameters.

03

Batch by job type, not by product

If you are generating 5 product videos, trigger all 5 jobs in parallel before polling any of them. Adly processes jobs concurrently; sequential triggering wastes wall-clock time.

For a deeper look at controlling the full creative workflow from an LLM interface, including job approval and copy generation, read the guide on how to use MCP to control marketing tools from Claude. For the specific UGC video formats Adly exposes through MCP, see the UGC AI video generator feature page.

Scene Studio output

Multi-scene commercials via the same MCP connection

Set job_type to "video_scenes" and pass a scene plan array to generate a full multi-scene commercial through Scene Studio. Hook, product reveal, proof, and CTA across up to 6 scenes, up to about 60 seconds total. The output URL schema is identical, so your Postiz mapping code does not change.

This is the same engine used for longer-form ad creative. One MCP connection covers UGC talking-head videos, animated product ads, and multi-scene commercials.

Scene Studio output, job #14651. Multi-scene with crossfade transitions.

Your pipeline is one MCP connection away from real video.

Generate your Adly MCP key, add the server block to your Claude config, and your next Postiz post can have a brand-aware video asset without a single manual step.

Get started with Adly Run Adly from Claude