What this covers
The pain: you grab a generic 100-point technical SEO checklist, half the items do not apply to your stack (no AMP, no Magento, no jQuery plugin), and the other half are vague enough that you cannot tell if you have done them. This workflow uses AI to generate a 30 to 60 item checklist specifically for your stack — Astro / Next / Hugo / WordPress — with priority labels and a verification command for every item, in about 20 minutes. The output is a markdown checklist you commit to the repo and refresh quarterly.
Who this is for
Indie devs running a content site, small SEO teams supporting a single product, and ICs onboarding to a new project who need to assess the SEO baseline. Especially useful when handing the site to a contractor: a stack-specific checklist with verification commands is the artifact the contractor actually uses, vs the generic PDF nobody reads.
When to reach for it
Before a launch, after a stack migration (Next 13 to 15, WordPress to Astro, anything that touches routing), when onboarding a new SEO contractor, or after a noticeable traffic dip where you cannot tell if the cause is technical or content. Also when Google rolls out an algorithm or schema update — refresh the checklist to capture what changed.
Before you start
- Inventory your stack precisely: framework + version, hosting provider, render mode (SSR / SSG / ISR / hybrid), CDN, image optimization, plugins, CMS, i18n setup. Each affects which checklist items apply.
- Decide the audience for the checklist: yourself (terse), contractor (with context), or stakeholder (with rationale). Each calls for a different verbosity.
- Pick a baseline site to compare against — an existing site on the same stack with known-good SEO. AI uses it as an anchor.
- Have one technical SEO blog post you trust as a sanity reference. You will use it to spot AI hallucinations like “set X-Robots-Tag to canonical”.
Step by step
- Describe your stack to AI in one paragraph. Include framework + version, hosting, render mode, CDN, image strategy, sitemap source, i18n approach. The more specific, the better the output.
- Ask AI to generate the checklist grouped into six sections: crawl, render, performance, schema, internationalization, monitoring. Six is the right number — fewer misses categories, more produces overlap.
- Ask AI to mark each item P0 (blocks launch or fixes a live regression), P1 (within 30 days), or P2 (nice to have). Force the model to use this rubric explicitly.
- Ask AI to add a verification command, URL pattern, or screenshot target for each item. Examples:
curl -I https://example.com/sitemap.xml,view-source: search for canonical, Lighthouse Performance score over 90. - Walk through every P0 item and ask: does this make sense for my specific stack? Astro SSG does not need ISR cache headers; SSR Next does. Remove inapplicable items.
- Save as your project’s
SEO_CHECKLIST.mdwith the date and a one-line stack snapshot. Commit it. Future you needs to know which version of the stack the checklist was for. - Re-generate quarterly. Search Console signals shift, the framework releases new features, schema vocabularies evolve.
First-run exercise
Run this on a single page first, not the whole site. Pick the homepage or your highest-traffic article. Generate the checklist, then verify every P0 item against that one page. Most teams find AI produces 6 to 10 items that do not apply (the page is SSG but AI assumed SSR) and 3 to 5 verification commands that are wrong (querying a property that does not exist). Mark these in the checklist file so the next quarterly refresh does not repeat them. For the second run, refine your stack description with what you learned, then regenerate.
Quality check
- Every item has a verification command, URL, or screenshot test. If it cannot be verified, it cannot be claimed done.
- Every P0 item is something you can actually fix in your codebase, not someone else’s. Hosting-provider items go in their own section.
- No item references a feature that does not exist in your framework version. “Use Next.js dynamic OG image route” only works in Next 13+.
- No duplicate items across sections. AI repeats canonical-tag-related items in both crawl and render.
- The checklist runs in under 60 minutes for someone familiar with the stack. If it takes longer, split into a P0-only quick check and a P1+P2 deep audit.
How to reuse this workflow
- Commit
SEO_CHECKLIST.mdto the repo. Treat it as part of CI: pre-launch checks run a subset of the verification commands automatically. - Maintain one checklist per public domain. A staging site has different rules (noindex) than production.
- Keep a
seo-misses/log: every time a real SEO issue slips into prod, add an item to the checklist that would have caught it. - Diff each quarterly regeneration against the previous one. The diff tells you what Google or the framework changed.
- Share the prompt with one peer running a different stack. Their feedback exposes which items are stack-specific vs universal.
Recommended workflow
Detailed stack description → AI generates grouped checklist (crawl / render / performance / schema / i18n / monitoring) → priority labels (P0 / P1 / P2) → verification commands per item → manual prune of inapplicable items → save as SEO_CHECKLIST.md in repo → quarterly refresh.
Common mistakes
- Starting from a generic checklist instead of generating one for your stack — you spend more time deleting than verifying.
- Skipping the verification command field. Without it, items rot — nobody can prove they are still done six months later.
- Treating the checklist as permanent. Google and frameworks ship changes; a 12-month-old checklist quietly misses real issues.
- Generating one giant checklist for multiple sites. Stack differences make this worse than per-site checklists.
- Letting AI invent commands. AI confidently produces
curl --google-botflags that do not exist. Verify every command runs. - Putting
update CMSorimprove contentas items. The checklist is technical SEO; content workflows live elsewhere.
FAQ
- How many items typically?: 30 to 60 after pruning, vs the 80 to 150 in generic templates. Larger is suspicious — usually means duplication or scope creep into content SEO.
- Astro-specific items?: Hydration boundaries (avoid client:load on above-fold static blocks), prerender mode, sitemap integration via
@astrojs/sitemap, hreflang via content collections. For the sitemap line, pair the check with a Codex sitemap review tutorial. - What about Next.js?: Render mode per route (SSG vs SSR vs ISR), metadata API usage, dynamic OG image route, middleware for redirects, ISR revalidation policy.
- How do I handle WordPress plugins?: List the active plugins in your stack description; AI will fold plugin-specific items in (e.g., Yoast canonical conflicts, RankMath schema duplicates).
- Should AI run the checklist for me?: AI can draft and verify; the verification command is what you actually run. Do not trust AI’s self-report that an item passes.
- Can I use this for non-content sites like SaaS landing pages?: Yes. The categories stay the same; the P0 items shift toward render mode and structured data for product schema.