Auto-Execute

Summary#

Auto-Execute takes one content topic and runs four tools in sequence — SEO Brief, Article Outline, Article Writer, Content Optimizer — then saves the finished article to your My Content library as a Draft. You watch each stage tick over as it runs. It costs 4 credits.

Purpose#

Producing one article normally means four separate runs and four handoffs: research the topic, structure it, write it, then check what you missed. Most of that time is spent moving output from one screen to the next, and the handoffs are where the work stalls.

Auto-Execute exists to remove the handoffs. The decision it supports is simple: do I want a finished first draft, or do I want the research kit? If you want the kit, use Content Engine. If you want a draft you can edit, use this.

Overview#

You enter a topic. Four stages run one after another, and each one feeds the next.

Stage 3 receives the brief's target keyword and the outline's structure as context, so the writer works from a real plan instead of guessing. Stage 4 scores the finished draft against that same brief.

Each stage retries once automatically on a transient failure. When the run finishes, the article is written into My Content with the tag auto-execute and the status draft. Nothing is published anywhere.

Benefits#

• One input produces a full draft, not a research kit you still have to write from.
• Every stage is visible while it runs, so a slow stage is never a mystery.
• Each stage retries itself once, and a failed optimizer does not lose the article.
• The finished draft saves itself, so a closed tab does not cost you the work.
• Five export formats plus the standard result exports are on the finished card.

Use Cases#

Filling an empty content calendar. You have ten topics and no writing time; each run turns one of them into an editable draft.

Getting past the blank page. You know the topic but not the shape — the draft arrives already structured, so editing replaces starting.

Testing whether a topic is worth a real article. A cheap first draft tells you how much there is to say before you commission a writer.

Producing a consistent baseline. Every draft follows the same brief-to-optimizer path, so a team's output has one structure rather than five.

Briefing a writer with something concrete. Hand over the draft and the optimizer's edit list rather than a blank brief.

Requirements#

• A signed-in account.
• A paid plan. All four stages cost credits, and any cost above zero is blocked on Free.
• At least 4 credits of monthly allowance available.
• A topic. This tool takes a subject, not a domain or a URL.

Permissions#

Cost#

4 credits per run — 1 credit for each of the four stages. The Quota Cost card above the input reads 4 / 4 credits per draft.

The run button is labelled Run Auto-Execute · 4 reports, matching the 4-credit charge the Quota Cost card shows.

A stage that fails after being charged is still charged, because the credit is taken when the call is accepted. A stage that runs from cached provider data still charges its credit — see Result caching and freshness and How credits work.

Navigation Path#

Dashboard → Get Started → Auto-Execute

Inputs#

Step-by-Step Guide#

1. Open Dashboard → Get Started → Auto-Execute.
2. Type your topic into the field marked content topic (e.g. email marketing).
3. Click Run Auto-Execute · 4 reports. The button changes to Running… do not close.
4. Watch the Auto-Execute pipeline card. Each of the four stage rows moves from queued to running… to done.
5. Expect one to three minutes end to end. Stage 3 is the slowest and says so: generating 1500+ words — usually 60-90s, please wait.
6. Read the finished card, starting with the score circle and the four tiles.
7. Edit or export from the buttons on the card. The draft is already saved.

Reading the Results#

The pipeline card. While the run is live, four rows report progress: Stage 1 — SEO Brief, Stage 2 — Article Outline, Stage 3 — Article Writer, Stage 4 — Content Optimizer. Each carries a hint that becomes a result when the stage finishes — Stage 1 replaces target keyword, intent, outline spec with the keyword it chose and the number of secondary keywords, Stage 2 with the count of H2 sections and title options, Stage 3 with the word count and estimated reading time, Stage 4 with Score: N/100.

A row reading retrying… (attempt 2 of 2) is normal — a transient timeout is retried once after a short pause, and the run usually completes.

The completion banner. AUTO-EXECUTE COMPLETE, the elapsed time and 4 stages, then a score circle on the right.

The score circle. A single 0-100 figure combining the writer's own SEO score and the optimizer's coverage score, with a word beneath it:

The title and meta description sit under the banner. Treat both as candidates, not finals — the title is the one thing worth rewriting by hand every time.

The four tiles.

Publish & save. Three buttons. Save to My Content stores the draft again (the run already saved it) and confirms with Saved to My Content as a Draft. Open in Article Generator copies the draft to your clipboard and opens AI Article Generator to continue it. Optimize Further does the same into Content Optimizer.

Export. Copy HTML, Copy Markdown, Word (.doc), PDF, Download HTML. Markdown is the one to use for most CMS editors — it is built from the clean article body, without the on-screen chrome. The Word and PDF exports strip the same chrome.

Generated Article · Magazine View. The article body itself. Read it as a first draft: the structure will be sound, the facts need checking, and the voice needs your hand. Nothing here has been published anywhere.

Frequently Asked Questions. The FAQ block the writer produced, with a question count badge. These map directly onto FAQ schema, which is why the next block exists.

Internal Linking Suggestions. Anchor text and the topic each link should point at, when the writer returned any. Pair this with Internal Linking Engine to find the real pages to point at.

JSON-LD Schema Markup Preview. A collapsed panel holding Article schema, plus FAQPage schema when there is an FAQ. The badge reads ARTICLE + FAQ or ARTICLE. Copy schema JSON-LD puts it on the clipboard ready to paste into the page head.

What to do with the whole thing. Fix the title. Check every factual claim and statistic. Replace the generic examples with your own. Then run the draft through Pre-Publish Trust Checks before it goes live.

Examples#

Screenshots#

Tips#

• Keep the tab open. The button says Running… do not close for a reason: the pipeline runs in your browser, and closing the tab abandons the stages that have not finished.
• Two or three words in the topic field gives the brief enough to work with and enough room to shape.
• If Stage 3 fails, use Retry just the article_writer stage first. It is cheaper in time than re-running everything.
• Run it once on a topic you know well. That is the fastest way to calibrate how much editing your drafts will need.
• The draft is saved before you touch anything, so you can navigate away and come back to it in My Content.

Best Practices#

• Treat the output as a first draft by a competent stranger. Check the facts, own the voice, keep the structure.
• Rewrite the title. It is the highest-leverage edit on the page and the one the model is weakest at.
• Add your own examples, screenshots and data. That is what makes the page worth ranking.
• Run Pre-Publish Trust Checks on the finished text before publishing.
• Use Content Engine instead when you want the research and intend to write the article yourself.

Common Mistakes#

• Entering a domain instead of a topic. The chain writes an article; a domain produces a confused brief and everything downstream inherits it.
• Publishing unedited. The draft is a starting point, and the score is about coverage, not accuracy.
• Closing the tab mid-run and expecting the article to be waiting.
• Re-running the same topic to "get a better one" instead of retrying the one stage that failed. A full re-run costs 4 credits again.
• Reading Rough draft as a bug. It usually means the topic was too broad.

Limitations#

• Four fixed stages, one topic per run, no way to change the chain.
• Stage time budgets are hard: 45, 50, 110 and 60 seconds. A stage that exceeds its budget is retried once and then fails.
• The article is short-form by design — roughly 1,500 words, not a 4,000-word guide.
• The pipeline runs in your browser. It does not continue in the background.
• There is no scheduling; every run is manual.
• The draft is saved to My Content only. It is never pushed to a CMS.
• The country and date-range pickers are sent with each stage but do not meaningfully change the article.
• Saved results are kept for 90 days.

Troubleshooting#

FAQs#

How much does it really cost? 4 credits — one for each stage. The button label reading 6 reports is out of date; the Quota Cost card showing 4 is correct. See How credits work.

Is the article published anywhere? No. It is saved to My Content as a Draft. The trust line under the hero says it plainly: saved to My Content, never published for you.

Why did it not ask me "you already ran this"? Auto-Execute marks itself as a workflow run, which suppresses the per-tool saved-result prompt so the chain does not stall waiting for a click. See How to run a tool.

Can I change which tools run, or add a fifth stage? No. The chain is fixed at brief, outline, writer, optimizer.

How is this different from the AI Article Generator? AI Article Generator is a set of tabs you drive one at a time and gives you control at each step. Auto-Execute runs the whole sequence unattended and hands back the finished draft.

Can I run it on a schedule? No. Every run is manual. Scheduled work lives in the automation tools instead.

Where does the draft go if I close the tab after it finishes? It is already in My Content, tagged auto-execute with the status Draft.