Automate your entire content production with the DuupFlow API
What if your whole production chain ran without you touching it? A Google Drive folder where you drop your raw clips, an AI that sorts them, DuupFlow generating the unique variants, and automatic per-account dispatch with a ready-to-publish schedule. That's exactly what the DuupFlow API key makes possible.
In this guide we break down the full pipeline shown in the video, and above all: we give you the exact prompt to copy-paste into Claude Code so it builds the whole automation for you — even if you don't code.
Full pipeline demo on video.
The 5-step pipeline
The idea: a linear flow that starts from your raw content and outputs unique variants, sorted and scheduled. Claude Code builds it once — then a cron re-runs it on its own, and you never touch it again.
The API key: the engine of the pipeline
At the heart of the flow is the DuupFlow API key. It lets you call DuupFlow from any tool — Make, n8n, a script, or Claude Code — without using the interface. One key unlocks every endpoint: image and video duplication, compression, AI-signature masking.
- Base:
https://www.duupflow.com/api/v1 - Authentication: header
Authorization: Bearer dflw_live_… - Video (async): you start a duplication job, then poll its status until the files are ready.
- Pro plan only — the key is generated in your dashboard, API tab.
The 5 steps in detail
1 · Drive — detect new clips
The script watches a “source” Google Drive folder. On each run it picks up the video files not yet processed and leaves the rest alone (thanks to a local state, it never processes the same clip twice).
2 · AI sort — filter & categorize
Each clip is classified automatically: which account / theme it belongs to, how many variants to make. Files that are too short or in the wrong format are skipped.
3 · DuupFlow — 1 clip → N variants
The core. For each clip, the script calls the DuupFlow API to generate N unique variants, waits for processing to finish, then downloads the files. Each variant is perceived as distinct — exactly what the algorithms require.
4 · Dispatch — sort by account
The variants are re-uploaded automatically into the right Google Drive subfolder: one folder per account, ready to feed your posts.
5 · Ready — schedule + auto log
Finally, the script generates a publishing schedule (which variant, which account, when) and a log of each run. You open your schedule, you publish — the rest is already done.
The prompt to copy-paste into Claude Code
Here's the full brief. Copy it, paste it into Claude Code, replace the API key and Drive folder ids, and let it build the pipeline. It will propose the architecture then code everything.
You are my automation engineer. Build a pipeline that runs entirely on its own and reposts my content as several unique variants, starting from my Google Drive, using the DuupFlow API.
CONTEXT
- I run several accounts (social media / OFM). Each clip must be published as several unique variants (one per account) so it isn't penalized for "unoriginal content".
- I have a DuupFlow API key (Pro plan): dflw_live_XXXXXXXXXXXX
- API base: https://www.duupflow.com/api/v1
- Auth: header Authorization: Bearer dflw_live_XXXXXXXXXXXX
THE PIPELINE (5 steps), run via a cron every hour:
1) DRIVE — Detect new clips
- Watch a SOURCE Google Drive folder (id: PUT_YOUR_FOLDER_ID_HERE).
- Pick up new video files (mp4/mov) not yet processed.
- Keep a local state (processed.json) of already-done ids so nothing is processed twice.
2) AI SORT — Filter & categorize
- For each new clip, infer the category / target account (filename, tag, or a quick analysis).
- Skip files that are too short, corrupted, or in the wrong format.
3) DUUPFLOW — Duplicate (1 clip → N variants)
- For each clip, start a duplication (N = number of accounts):
POST /videos/duplicate (multipart/form-data)
file=@clip.mp4 count=N packs=visual,motion,metadata_technical
-> 202 response: { "job_id": "...", "status": "queued" }
- Poll GET /jobs/{job_id} every 5s until "status": "completed".
- Download each variant from result.files[].url (with the Authorization header). Available for 16h.
4) DISPATCH — Sort by account
- Re-upload each variant into the matching account's Google Drive subfolder (one folder per account).
5) READY — Schedule + auto log
- Generate a publishing schedule (CSV or Google Sheet): which variant, which account, which date/time.
- Write a log per run (date, source clip, number of variants, accounts, status).
CONSTRAINTS
- Idempotent: safe to re-run without duplicates thanks to processed.json.
- Robust: retry API calls, handle job timeouts.
- Secrets in a .env (API key, Drive folder ids) — never hardcoded.
- Add the cron (crontab or GitHub Actions) that runs the script every hour.
Stack: Node.js or Python, your call. Start by proposing the file structure, then code everything.💡 Tip: keep your API key secret — paste it only into your local .env file, never into a public repository.
Get started
You need two things: a Pro plan to generate your API key, and five minutes to run the prompt in Claude Code. Once the pipeline is in place, your production runs on autopilot.
Frequently asked questions
Do I need to know how to code to set this up?
No. The idea: you copy the prompt below, paste it into Claude Code, and it builds the pipeline for you. All you do is fill in your DuupFlow API key and your Google Drive folder ids.
What exactly does the DuupFlow API key let me do?
It lets you call DuupFlow from any tool (Make, n8n, a script, Claude Code) to duplicate images and videos into unique variants, compress, or mask the AI signature — without using the interface. One key unlocks every endpoint. It's available on the Pro plan.
How many variants can I generate per video?
Up to 10 variants per call on the video endpoint. You choose the number with the count parameter, and the transformation packs (visual, motion, metadata, magic pixel) with the packs parameter.
Does the pipeline really run on its own?
Yes. Once built, a cron runs it at the interval you choose (e.g. every hour). It detects new clips on the Drive, duplicates them via the API, sorts the variants by account and updates your schedule — hands-off.
Where do I get my API key?
In your DuupFlow dashboard, API (Developers) tab. Generate a key, copy it (it's shown only once) and paste it into your .env.