Meta tags example: generate only the tags this page can support
A practical meta-tags prompt that produces title, description, canonical, robots and social tags from verified page facts without invented markup.
- Works in
- ChatGPT, Claude, Gemini
- You need
- Page purpose and copy · Canonical URL · Indexing decision · Social image details
- Written for
- meta tags example
The correct meta-tag block depends on decisions outside the markup. A model cannot decide your preferred canonical or whether a URL belongs in search. This prompt makes those inputs explicit and omits what it cannot know.
Preview the finished title and description with the SERP snippet preview, then inspect the public HTML for duplicates.
Produce the HTML head tags for this one page, using only supplied facts.
First return a decision table: Tag | Include? | Supplied value | Missing fact.
Consider title, meta description, canonical, robots, viewport, charset,
Open Graph title/description/url/type/image/image alt, and Twitter card.
Then output one copyable HTML block. Never invent a canonical URL, social
image, locale, author, publication date, noindex decision or card handle.
Write [NEEDS VALUE] in the decision table and omit that tag from the code.
Do not output keywords meta tags. Do not use both name="robots" noindex and
language claiming the page should rank without flagging the conflict.
Keep title and description faithful to the visible page. Use absolute URLs.
Escape HTML characters. After the code, report conflicts with the supplied
existing head and list what must be checked in rendered HTML.
Page purpose and visible copy: [PAGE]
Canonical public URL: [CANONICAL]
Indexing decision: [INDEXING]
Social image URL, dimensions and alt: [IMAGE]
Existing head: [EXISTING]What to change
Everything in square brackets is yours to replace. Nothing else needs editing.
[PAGE]- Page type, H1, opening and verified differentiator.
[CANONICAL]- The final preferred absolute URL, decided by the site owner.
[INDEXING]- Index/follow, noindex/follow or another deliberate policy.
[IMAGE]- Absolute image URL, dimensions and meaningful alternative, or none.
[EXISTING]- Current rendered head so duplicate tags can be identified.
How to run it
- 01Decide canonical and indexing first
Those are site decisions, not copy the model should infer.
- 02Install one source of truth
Update the template or SEO field that owns each tag instead of pasting duplicates into another layer.
- 03Inspect rendered HTML
Confirm one title, description, canonical and robots directive appear in the final response.
Questions people ask
What are meta tags?
Meta tags and related head elements describe a page to browsers, search engines and social platforms. Common examples include the description, robots directive, canonical link and Open Graph properties.
Is the meta keywords tag useful for SEO?
No. Modern search engines do not use the keywords meta tag as a ranking input, so this prompt deliberately excludes it.
Can a page have two meta descriptions?
HTML may contain them, but competing descriptions create ambiguity and usually indicate two templates or plugins are trying to own the same field. Keep one intentional description.
Unsubscribe in one click. We never pass your address on.