Swiss Knife SEO

Home / How it works

One loop, run until the fix is verified

Most tools stop at a list of problems. Swiss Knife SEO runs a single loop that carries each finding all the way through: find it, plan the fix from your own crawl data, publish the change, and verify it on the next crawl. Here is exactly how that works.

The core loop

Four steps, and the last one closes the circle

Every finding moves through the same four stages. Nothing is left as a number in a spreadsheet, and nothing is marked done until a fresh crawl agrees it is done.

1

Find

The crawler renders your pages with a real browser, the same way a search engine does, then scores the site from 0 to 100. Findings are ranked by impact and by template reach, so a single meta rule that touches 400 pages sits above a one-off typo. You see the exact URLs behind every number, not a bucket average.

2

Plan

Ask AI for a fix plan and it is written from your own crawl data: the affected URLs, the crawler messages, and the page titles behind the issue. That means the plan matches your CMS and your templates instead of handing back generic advice you still have to translate.

3

Publish

Take a copy-paste patch for titles, metas, robots directives, canonicals, or security headers and drop it into your stack. Or publish the change at the CDN edge, where the response is rewritten in flight without touching your CMS, with one-click revert if you change your mind.

4

Verify

The next crawl rechecks the exact URLs against the rule the fix had to satisfy and marks the issue verified, noting the crawl that confirmed it. Owners, due dates, and SLA timers travel with each task, so work is closed on evidence, not on a promise.

Renders JavaScript Ranked by impact and reach Grounded in your URLs Copy-paste or edge publish One-click revert Auto re-check Owners and SLAs
The differentiator

Why a body link and a footer link are not equal

A normal crawl export counts internal links and stops there. Ours reads where each link sits and computes a position-weighted internal PageRank for every page, because a link in the middle of a paragraph carries far more weight than the same link repeated in the footer of every template.

Orphan pages

Pages that no other page links to in the body sit outside the flow of internal equity. A raw link count can miss them entirely because a boilerplate menu technically points at them. Position weighting exposes them as the starved pages they are.

Boilerplate-only pages

Some pages are reached only through nav and footer links that repeat sitewide. They look linked in a flat export, but they earn almost no real equity. We flag them so you can add a genuine in-body link from a relevant page.

Low-equity money pages

Your category and product pages should be strong, yet many quietly sit near the bottom of the internal equity map. The report surfaces the high-value pages that are underlinked and names the source pages and anchors that would fix it.

The practical payoff: the same crawl that finds these gaps also produces a ranked list of internal links to add, each with the source page and the anchor text to use. It is the difference between knowing a page is weak and knowing the exact three links that would strengthen it.

Read the score honestly

A low first score is a baseline, not a verdict

The health score is computed from weighted issues per crawled page, so it cannot be flattered by crawling fewer URLs. A real site with real history usually starts lower than people expect.

A first crawl commonly lands somewhere in the 20s to 40s out of 100. That is normal, and it is the baseline you improve from, not a judgment on your work. Because findings cluster by template, clearing one rule often lifts hundreds of pages at once, so the number moves in real steps. The score that matters is the trend across crawls, and every point on it links straight back to the pages behind it.

  • Scored per crawled page, so it stays honest as the site grows
  • Thematic sub-scores for crawlability, linking, on-page, and security show where the number comes from
  • Every point clicks through to the exact URLs, so nothing hides in an average
  • Template-level fixes move the score in visible jumps between crawls
Common questions

Before you run the loop

Do I need to install anything to run the loop?
No. Add your URL and the crawl runs in the cloud with a real browser, so you do not install or update a desktop app. Publishing a copy-paste patch means pasting a snippet into your own CMS or server, and edge publishing rewrites the response at the CDN without touching your CMS at all. The only thing you connect, and only if you want traffic context, is Search Console and GA4.
What does the first crawl usually score?
A first crawl commonly lands in the 20s to 40s out of 100. That is the baseline to improve, not a verdict on your site. The score is computed from weighted issues per crawled page, so it cannot be flattered by crawling less, and the number that matters is the trend across crawls as you clear findings by template.
How does a fix get marked verified?
Every fix keeps the exact URLs and the rule it needs to satisfy. On the next crawl those URLs are rechecked against that rule. When they pass, the issue is marked verified with the crawl number that confirmed it. Owners, due dates, and SLA timers travel with each task, so nothing is closed on a promise.

See the loop on your own site

Run a crawl, get findings ranked by impact, and watch the first fix carry through to verified. The free plan includes the full report suite on up to 100 URLs.