This is the page to bookmark. Whenever you’re looking at a number or color in MAEL and wondering “is that good?”, come back here.

​
Run and step statuses

Every workflow run and every step inside it has a status badge with a color and an icon — never color alone, so it’s always readable.

​
Quality scores (SEO score, Readability score)

Content articles carry two scores, shown as ring gauges (0–100) wherever an article appears — the Content board, Library, and an article’s detail view. SEO score reflects keyword usage, structure, and on-page optimization. Readability score reflects how easy the text is to read (sentence length, clarity, structure). Both are produced by the AI review agents that ran as part of your content workflow, not typed in by a person.

​
The content evaluation gate (why some articles need review)

Every article that goes through the full content pipeline passes through an automatic quality check before it can reach a human for approval. It scores nine dimensions — including medical accuracy and the risk that the text reads as AI-generated or contains a fabricated claim — and:
  • Passes → the article moves to the approval queue (a human still has to say yes — see Take Action)
  • Fails → the article’s status becomes review_required instead — it’s held back from the approval queue until someone looks at why it failed
review_required is not a bug and not a dead end. It means the automatic quality check caught something worth a second look before it goes any further — that’s the system working as intended.

​
Errors vs. warnings

  • A red error box (on a run’s detail page, right under the status badge) means the run — or the step — actually failed. It includes a short error code and a human-readable message.
  • A yellow status (like review_required or waiting_approval) is not an error — it’s a deliberate pause, waiting on you.
If you see a red error, the message itself usually tells you what went wrong. For the common ones and what to do, see Troubleshooting.

​
Cost and tokens

Every run shows Cost (USD) and Tokens (a unit of AI usage — see the Glossary). These aren’t something to “fix” — they’re informational, so you know what a workflow actually costs. If a number looks unexpectedly high, compare it against similar past runs on the Workflows list before assuming something’s wrong.

​
Analytics numbers (clicks, impressions, CTR, position)

On the Analytics page: These come directly from your connected Google Search Console / GA4 properties — MAEL doesn’t modify or estimate them.

​
What to do with each kind of result

Open it and look at what it produced (see the relevant SEO Workflow’s page for what “output” means for that specific workflow). For content, that usually means reviewing it in Take Action before it can publish.
Read the error message on the run page. Check Troubleshooting for the common causes. If it’s not obvious, note the run ID and error code and pass it to your administrator.
Don’t approve it as-is. Either reject it with a comment explaining what to fix (see Take Action), or treat it as a candidate for the content-refresh workflow later.
Read why (the evaluation gate’s findings) before deciding — this status exists specifically so a low-confidence or risky article doesn’t sail straight to your approval queue unlabeled.
Each issue is a specific, real finding (a crawl error, a slow page, a missing structured-data field) — see Technical SEO Audit for how to prioritize them.

Next: Take Action