seo-drift

Overview of seo-drift skill

What seo-drift does

The seo-drift skill monitors SEO-critical page elements over time so you can catch regressions after a deploy, CMS edit, or template change. It is most useful when you need a baseline, a comparison, or a change history for a specific URL and want a fast answer to “did anything break?” rather than a vague SEO audit.

Who should install it

Install the seo-drift skill if you manage pages where technical on-page SEO matters: titles, canonicals, robots directives, headings, schema, and social metadata. It is a strong fit for SEO teams, content ops, developers shipping marketing pages, and anyone responsible for preventing silent SEO regressions.

Why it is different

Unlike a generic prompt, seo-drift is built around repeatable compare logic and severity levels. That means it is aimed at finding meaningful changes, not just summarizing a page. The skill is especially valuable when you care about drift detection across versions, not one-time page analysis.

How to Use seo-drift skill

Install and start with the right files

Use the install path shown in the repo and then read skills/seo-drift/SKILL.md first. For implementation details, open references/comparison-rules.md next, since it defines what the seo-drift skill actually treats as a regression. If you are adapting the skill to your own workflow, inspect any command examples, field mappings, and rule thresholds before you rely on its output.

Give the skill a narrow, testable goal

The best seo-drift usage starts with one URL and one question. Good inputs look like: “Compare this product page against its last baseline and flag any SEO drift that could affect indexing.” Better still: include the page type, the expected canonical behavior, and whether the page is allowed to change titles during experiments. The skill works best when you tell it what “normal” looks like.

Prompt patterns that work well

For a baseline, ask for a clean snapshot of a known-good page and specify what must be preserved. For a comparison, provide the current URL plus the baseline reference and mention whether you care most about critical regressions or a full diff. For history, ask for the last meaningful changes and the likely SEO impact. This makes seo-drift usage more actionable than a plain “check this page” prompt.

Read outputs in severity order

Use the comparison results in this order: critical issues first, then structural changes, then lower-risk drift. That matters because seo-drift is designed to prioritize problems that can affect indexing or rankings quickly, such as missing canonicals, added noindex directives, removed schema, or heading loss. If your workflow is time-boxed, fix or verify the top severity items before reviewing cosmetic changes.

seo-drift skill FAQ

Is seo-drift only for technical SEO?

No. The seo-drift skill is not a full audit suite, but it is broader than “technical only.” It tracks the page elements that often drive content and indexing stability, which makes it useful for SEO Content checks after publishing or editing.

When should I not use seo-drift?

Do not use it when you need keyword research, content briefs, backlink analysis, or a broad site crawl. seo-drift is for page-level regression detection. If your question is “what should this page rank for?” or “how should I rewrite this article?”, a different workflow will be more useful.

Is it beginner-friendly?

Yes, if you can provide a URL and a clear comparison goal. The main failure mode is under-specifying what baseline or version should be treated as authoritative. A beginner can get good results quickly, but the output improves a lot when the prompt says what changed, what should not change, and what outcome matters.

How is seo-drift different from a normal prompt?

A normal prompt can describe SEO checks, but the seo-drift skill encodes comparison behavior and regression logic. That reduces guesswork when you need repeatable reviews after deployments. In practice, that makes seo-drift better for teams that need a consistent seo-drift guide rather than an ad hoc one-off prompt.

How to Improve seo-drift skill

Specify the baseline and the acceptable drift

The biggest quality gain comes from telling seo-drift what baseline is authoritative and what edits are allowed. For example, say whether title changes are expected for seasonal campaigns, whether canonicals must remain fixed, or whether H2 restructuring is acceptable. Without this context, the skill may flag intended edits as regressions or miss business-critical exceptions.

Add page context that affects interpretation

Strong inputs include page type, template name, locale, deployment window, and whether the page is indexable by design. That context helps the seo-drift skill distinguish real SEO damage from normal content evolution. For SEO Content pages, also mention whether the page is editorial, product-led, or programmatic, because acceptable drift differs by format.

Ask for a decision, not just a diff

If you want better output, ask for “what changed, why it matters, and what to do next” instead of a raw list of differences. The skill is most valuable when it ranks issues by likely SEO impact and tells you which changes need immediate rollback, verification, or documentation. That turns the comparison into an operational check, not a report artifact.

Iterate with the first comparison

After the first run, refine the prompt based on false positives or missing checks. If headings changed intentionally, say so. If schema changes are expected, narrow the review to canonicals and robots directives. If a template rollout caused noise, ask seo-drift to focus on only the rules that matter for the page class. That is the fastest way to make the seo-drift install pay off in production.