Paste one of your AWS Builder Center articles. You'll get its score and a ranked list of what would raise it — each with the number of points it is worth. Where a scoring cap is the thing holding you back, that number is exact arithmetic, not a guess.
Nothing is stored and nothing is published. This page never adds an article to the public leaderboard.
Eight deterministic tools measure the article first — whether the code parses, whether the AWS APIs exist, whether the links resolve, whether there is any pasted command output. Four judges then score five dimensions, and the final number is a weighted sum:
| insight | 10 |
| aws service depth | 10 |
| substance | 9 |
| evidence | 9 |
| clarity | 8 |
| explanation depth | 7 |
| architecture quality | 7 |
| code quality | 7 |
| actionability | 7 |
| accuracy | 6 |
| sources | 6 |
| well architected | 5 |
| scope discipline | 3 |
| currency | 3 |
| cost awareness | 3 |
Each dimension is capped by what was measured, so a judge cannot award points for code to an article with none. A cap is usually the biggest single thing standing between a draft and a better score, which is why the recommendations lead with them. 70+ is READ, 40–69 is SKIM.