fixing-metadata
by ibelickThe fixing-metadata skill helps audit and repair HTML metadata for pages that need to be discoverable, indexable, and shareable. Use it for frontend pages, SEO templates, marketing pages, and shared layout defaults when titles, descriptions, canonicals, Open Graph tags, Twitter cards, favicons, JSON-LD, or robots directives need to be corrected.
This skill scores 74/100, which means it is worth listing: it gives agents a concrete metadata-fixing workflow with clear triggers and scoped actions, though users should expect some gaps in operational depth. For directory users, it looks installable for routine SEO/meta-tag tasks, but not yet as fully self-serve as a richer skill with examples, checks, or supporting assets.
- Clear triggerability for metadata work: title, description, canonical, Open Graph, Twitter cards, favicons, JSON-LD, and robots directives are explicitly named.
- Workflow is operationally useful: identify issues, prioritize critical duplication/indexing problems, align title/description/canonical/og:url, and keep diffs metadata-only.
- Good install decision value for common web SEO tasks: it covers new pages, shared SEO components, locale/canonical routing, and share previews.
- No support files, scripts, or references are included, so users must rely on the SKILL.md instructions alone.
- The hierarchy/table appears partially truncated in the excerpt, which weakens confidence in how comprehensive the priority rules and edge-case guidance are.
Overview of fixing-metadata skill
What fixing-metadata does
The fixing-metadata skill helps you audit and repair HTML metadata for pages that need to be discoverable, indexable, and shareable. It covers page titles, meta descriptions, canonical URLs, Open Graph tags, Twitter cards, favicons, JSON-LD, and robots directives.
Who should use it
Use this fixing-metadata skill if you are shipping frontend pages, SEO templates, marketing pages, or shared layout defaults and need metadata that is correct rather than merely present. It is especially useful for Frontend Development workflows where metadata lives in app layouts, page components, or framework-specific head utilities.
Why it is different
The skill is built around priority order: fix duplicates and indexing problems first, then align title, description, canonical, and social tags, and only after that polish icons or structured data. That makes the fixing-metadata guide practical for adoption because it emphasizes decisions that affect crawlability and previews, not just tag completeness.
How to Use fixing-metadata skill
Install and locate the core instructions
For fixing-metadata install, add the skill to your skills directory or workspace, then open SKILL.md first. This repository is intentionally compact, so the main value comes from following the workflow and rules in that file rather than searching for supporting folders that do not exist here.
Turn a rough task into a useful prompt
The best fixing-metadata usage starts with a concrete page or template scope. Instead of saying “fix SEO,” give the agent the page type, framework, and the metadata problem. For example: “Audit the product detail template in Next.js and fix title, description, canonical, og:url, Twitter card, and robots tags for duplicate indexable URLs.”
What input improves output quality
Provide the page list, target URLs, desired canonical behavior, and any brand or locale rules before asking for changes. The skill works best when you specify whether the goal is a new page, a template default, or a repair of broken metadata, because those situations require different tradeoffs.
Practical workflow and review points
Start with the affected pages, map each one to the metadata that must agree, then check the output on a real deployed URL rather than localhost. Review minimal diffs only: the skill is meant to constrain change to metadata, not refactor layout logic or unrelated frontend code.
fixing-metadata skill FAQ
Is fixing-metadata only for SEO work?
No. The fixing-metadata skill also helps with social preview quality, app icons, manifest settings, and structured data. SEO is the main use case, but the real job is making page identity consistent across search engines, browsers, and sharing platforms.
Do I need to be advanced to use it?
No. Beginners can use this fixing-metadata skill if they can name the page and describe the problem. The biggest success factor is giving a specific target, because vague requests tend to produce generic metadata that looks correct but does not fit the route or framework.
When should I not use this skill?
Do not use it if the issue is mainly content strategy, keyword research, or copywriting. This skill is for implementation and validation of metadata, not for deciding what the page should say at a marketing level.
How does it compare with a normal prompt?
A normal prompt may produce a one-off fix, but fixing-metadata adds a guided workflow for priority, consistency, and deployment-safe review. That matters when you need metadata changes that survive framework rendering, canonical routing, and share-card checks.
How to Improve fixing-metadata skill
Give the skill the exact metadata surface
The strongest fixing-metadata usage includes the page URL, route pattern, framework, and which tags already exist. If you can paste the current <head> output or the relevant layout component, the agent can focus on what is missing or inconsistent instead of guessing.
State the rules that matter most
The skill performs better when you define canonical logic, locale behavior, and indexing rules up front. For example, say whether query parameters should canonicalize to the clean path, whether locale variants need alternate URLs, and whether robots should allow or block specific templates.
Watch for the common failure modes
The most common problems are mismatched title and og:title, descriptions that are too generic, canonicals that point to the wrong route, and social tags that only work in local previews. If you want higher-quality output, ask the agent to verify agreement across all metadata sources before editing.
Iterate after the first pass
After the first fix, validate the deployed page as if you were a crawler and a social platform. If anything is still off, feed back the exact symptom, such as a duplicate title, incorrect preview image, or wrong canonical, so the next pass can stay narrowly scoped and preserve the rest of the frontend implementation.