SMF·SEOBILITY
Can a prompt replace Seobility?
SEO & marketing — technical audits and on-page optimization
Exhibit tracking slip
Verdict
Every audit tool produces findings and most of them stop there, which is why audits get exported to a spreadsheet and forgotten. Seobility's useful habit is treating a finding as a ticket with an owner and a state that the next crawl resolves automatically. Building that lifecycle — stable finding identity across crawls, so "the same problem" is recognised weeks later — is more subtle than the crawler itself, and it is what makes an audit tool something a team actually uses twice.
Exhibit A — The prompt
Received on31.07.2026Build a site auditor whose findings have a lifecycle across crawls.
Guard: require confirmation of ownership or permission before any crawl, recorded with the crawl.
Crawler: descriptive user agent, robots.txt honoured, configurable concurrency, delay and URL cap. Internal links only. Store status, headers, timing and body per URL. Persist crawls so any two can be compared.
Rules: an editable file of checks, each with an id, severity, explanation and remediation note. Cover indexability (noindex on a linked page, canonical to a redirect or off-site, redirect chains, blocked pages in the sitemap), content (missing or duplicate titles and descriptions, missing or multiple H1, thin pages, near-duplicate bodies), links (broken internal links, orphans, click depth, generic anchors), and hygiene (uncompressed responses, oversized images, missing alt text, mixed content).
Finding identity, which is the point of this build and the part that needs care. A finding is not "this rule fired on this URL in this crawl" — it is a problem that persists across crawls. Compute a stable identity from the rule id plus a normalised URL: strip the scheme, lowercase the host, drop trailing slashes and a configurable set of tracking parameters, and resolve the URL through the redirect chain recorded in the crawl. Without that normalisation, a site that adds a query parameter or flips to HTTPS produces a hundred "new" findings that are the same hundred problems.
Lifecycle: a finding is open, in progress, fixed, or won't-fix. A crawl in which the rule no longer fires for that identity moves an open finding to fixed and records the crawl that closed it. A finding that reappears after being fixed is reopened and marked as a regression — regressions are the most valuable signal an audit tool produces and every report should lead with them. A won't-fix finding stays suppressed until it regresses, at which point it surfaces again with its previous decision attached.
Work queue: findings ordered by severity then by page importance (click depth, and internal links pointing at the page), with an assignee, a note and a target date per finding. Bulk state changes across a filtered selection.
Crawl comparison: for any two crawls, what opened, what closed, what regressed, and how each rule's count moved.
Scheduling: recurring crawls with an alert when regressions appear or when the open count crosses a threshold.
Export: findings as CSV, a self-contained HTML report that opens with no server, and a full JSON dump.
Out of scope: JavaScript rendering, keyword or backlink data, crawling sites the user has not confirmed owning, and automatic changes to a live site.Opening prefills the prompt — press enter to run it.
Exhibit B — What you lose
- B.1 keyword and backlink monitoring bundled into the same subscription
- B.2 hosted crawling of large sites without occupying your own machine
- B.3 the client-facing report generation
- B.4 JavaScript rendering for sites that need it to be crawlable
Prior art
Exhibit C — Why people still pay: crawl scale, rule depth, and operational polish
Because fifty dollars buys the crawl, the keyword tracking and the backlink monitoring together, and the two datasets you cannot rebuild are the reason the bundle exists.
Questions
Can I import my Seobility findings?
Findings export as CSV and can be loaded as historical records, but they will not share identities with what this build computes, so the lifecycle starts fresh on the first crawl. Any won't-fix decisions are worth re-entering by hand.
Why is finding identity the hard part?
Because without stable identities, every crawl produces a fresh list and the tool cannot tell you what got fixed or what came back. Normalising the URL through redirects and stripping tracking parameters is unglamorous and it is the difference between a work queue and a wall of noise.
What does it cost to run?
Nothing but a small VPS if you want scheduled crawls, five to ten dollars a month, or nothing at all running it locally on demand. Crawl data is small; a site of ten thousand pages is a few hundred megabytes with bodies stored.
What is the one thing that does not survive the rebuild?
The bundle. Seobility's price covers the crawl plus keyword tracking plus backlink monitoring, and the second and third run on datasets nobody can rebuild. An audit tool alone answers a third of the questions the subscription was bought for.
Related tools
Receipt