internal-linking-optimizer
by aaron-he-zhuinternal-linking-optimizer is a planning skill for SEO content teams to improve site architecture, fix orphan pages, reduce crawl depth, and map stronger internal links with reusable templates and examples.
This skill scores 81/100, which means it is a solid directory listing candidate: agents get strong trigger coverage, substantial workflow guidance, and reusable output structure for internal linking work, though users should still expect some manual adaptation because the repo provides guidance and templates rather than executable tooling.
- Very triggerable: frontmatter includes many explicit phrases and multilingual triggers for internal linking, site architecture, orphan pages, and link equity use cases.
- Operationally rich: SKILL.md is substantial and supported by three focused reference docs covering architecture patterns, worked examples, and output templates.
- Better than a generic prompt for repeatable SEO analysis: it gives concrete link-planning frameworks, implementation steps, and structured deliverable formats.
- No install command or automation assets are provided, so adoption depends on reading the docs and applying the process manually in the agent workflow.
- Evidence points to strategy and recommendation generation, but not to built-in site crawling, data extraction, or direct CMS/tool integration.
Overview of internal-linking-optimizer skill
What internal-linking-optimizer actually does
internal-linking-optimizer is a planning and analysis skill for improving internal links across a content site. It is designed for SEO content teams, site owners, and AI-assisted editors who need a clearer link structure, fewer orphan pages, better crawl paths, and more intentional distribution of page authority.
Who should install this skill
This skill fits best if you already have multiple articles, service pages, or topic clusters and need a repeatable way to decide what should link where. It is especially useful for content-heavy sites where “site structure is messy,” pages have weak internal support, or pillar and cluster relationships are unclear.
Real job-to-be-done
Most users are not looking for theory. They want an actionable internal linking plan: which pages should become hubs, which pages are underlinked, where contextual links should be inserted, and how to reduce crawl depth without creating random, low-value links. That is the core value of the internal-linking-optimizer skill.
What makes this skill different from a generic prompt
The repository is stronger than a one-off “suggest some internal links” prompt because it includes architecture patterns, a worked example, and reusable output templates. Those references push the model toward structured deliverables such as cluster maps, link opportunity tables, and prioritized actions instead of loose SEO advice.
What to check before adopting
This skill is most valuable when you can provide a page list, URLs, topics, existing hierarchy, or sample content. If you only provide a homepage URL and ask for a full internal-linking strategy, the output will be generic. The skill helps most when paired with real site structure data.
How to Use internal-linking-optimizer skill
Install context and compatibility
The upstream skill declares compatibility with Claude Code ≥1.0, skills.sh marketplace, ClawHub marketplace, and the Vercel Labs skills ecosystem. It does not require system packages. Optional networked SEO tooling can help, but the skill can still be used with manually supplied site data.
Where to read first in the repository
Start with optimize/internal-linking-optimizer/SKILL.md for trigger conditions and workflow. Then read:
optimize/internal-linking-optimizer/references/link-architecture-patterns.mdoptimize/internal-linking-optimizer/references/linking-example.mdoptimize/internal-linking-optimizer/references/linking-templates.md
These three files matter because they show the intended output shape, not just the topic.
What inputs the skill needs
For strong internal-linking-optimizer usage, provide as many of these as possible:
- A list of important URLs
- Page types such as blog, category, service, feature, docs
- Primary topic or keyword per page
- Existing pillar pages or suspected hubs
- Pages with low traffic, no rankings, or no internal links
- Navigation constraints and pages you do not want overlinked
- Sample article text or excerpts where contextual links can be inserted
Best input format for reliable output
A simple table or bullet list works better than a vague request. Good inputs usually include URL, page purpose, target keyword, current parent topic, and priority. This gives the skill enough structure to recommend links that support both relevance and hierarchy.
Turn a rough ask into a strong prompt
Weak prompt:
“Improve internal links on my site.”
Stronger prompt:
“Use internal-linking-optimizer for SEO content. Analyze these 25 URLs, identify orphan or weakly connected pages, propose a hub-and-spoke structure, recommend inbound and outbound internal links, and prioritize quick wins that improve crawl depth and topical authority. Use tables for from page, to page, anchor, reason, and priority.”
Ask for the right deliverables
The skill is most useful when you request outputs that match its references:
- A current structure diagnosis
- A proposed architecture model
- A page-by-page link opportunity list
- Inbound links each priority page should receive
- Suggested anchor text variants
- A phased implementation plan
That keeps the response operational instead of abstract.
Use the architecture patterns deliberately
The repository includes concrete models such as hub-and-spoke topic clusters. Use them as decision frameworks, not as defaults. If your site is service-led, a strict content-cluster model may be less useful than a structure centered on commercial pages, support content, and conversion paths.
Use the worked example to shape output
references/linking-example.md shows the expected level of specificity: where on the page to add the link, what text to link, and which page should receive it. If your first result is too high-level, ask the model to mirror that worked-example format.
Use the templates when scaling
references/linking-templates.md is the practical advantage of this skill. It helps the model produce consistent reports across many pages, including cluster strategy, contextual opportunities, and link tables. For teams, this makes the internal-linking-optimizer guide easier to operationalize across writers and editors.
Suggested workflow for real sites
Use this order:
- Inventory pages and tag topics
- Identify hubs, clusters, and orphan pages
- Choose the target architecture
- Generate contextual links and inbound support plans
- Review anchors for naturalness and intent match
- Implement highest-impact links first
- Re-check whether key pages now sit closer to the main crawl path
What affects output quality most
The biggest quality lever is page-level context. If the model sees only titles, it can still map architecture, but contextual anchor suggestions will be weaker. If you provide excerpts or summaries, the skill can recommend links that fit naturally within actual paragraphs.
What this skill does not replace
internal-linking-optimizer install does not replace crawling software, analytics, or manual editorial review. It helps you generate and organize decisions. You still need to verify whether a recommended link belongs in the page’s real user journey and whether the anchor feels editorially natural.
internal-linking-optimizer skill FAQ
Is internal-linking-optimizer good for beginners
Yes, if you already understand your site’s basic pages. The references make the workflow more concrete than many SEO skills. Beginners may still need to supply a cleaner URL inventory than advanced users, because the skill is not a crawler by itself.
Is this only for blogs
No. It is a good fit for blogs, resource centers, documentation, SaaS marketing sites, and service businesses. The key requirement is that internal links matter to discovery, topical grouping, and authority flow across multiple pages.
When is this skill a poor fit
It is a weak fit for tiny sites with only a few pages, for single-landing-page projects, or for teams that want a fully automated crawl-and-implement system. If there is no meaningful content graph to improve, the skill will have little leverage.
How is it different from asking an AI for link ideas
A generic prompt often returns generic advice like “add relevant links.” The internal-linking-optimizer skill is more useful because the repository pushes toward architecture diagnosis, orphan-page handling, and reusable output structures that can be implemented by content teams.
Can it help with orphan pages
Yes. Orphan pages are one of the clearest trigger cases in the skill metadata. If you provide a page inventory and identify pages with no inbound internal links, the skill can propose which hubs, related articles, or navigational pages should support them.
Does it work for multilingual sites
Potentially, yes, because the trigger set is multilingual and the concept is language-agnostic. But output quality still depends on whether you provide page data in a structured way and whether you clarify cross-language linking rules or locale boundaries.
How to Improve internal-linking-optimizer skill
Give it a site map, not just a topic
The fastest way to improve internal-linking-optimizer for SEO Content is to provide a mini site map. Even a rough list of homepage > category > article relationships helps the model distinguish structural fixes from random contextual link suggestions.
Mark business priorities explicitly
Tell the skill which pages matter most: revenue pages, key pillar pages, high-converting resources, or strategic articles. Otherwise it may distribute attention too evenly instead of directing link equity toward pages that matter commercially.
Separate architecture work from copy-level work
Run two passes:
- Pass 1 for hierarchy, hubs, orphan pages, and crawl depth
- Pass 2 for paragraph-level contextual links and anchors
This reduces muddled outputs and produces clearer implementation tasks.
Provide content excerpts for anchor recommendations
If you want usable anchors, include the paragraph or section where the link could appear. Without that context, anchor suggestions are often accurate in topic but awkward in phrasing. With excerpts, the skill can propose links that fit the sentence naturally.
Ask for reasons, not just links
A better prompt asks the model to explain each recommendation with labels such as topical relevance, authority support, crawl path improvement, or user journey. Those reasons make it easier to filter weak ideas before implementation.
Watch for common failure modes
The most common problems are:
- Linking pages only because they share a keyword
- Overloading a page with too many links
- Ignoring commercial pages in favor of informational ones
- Suggesting anchors that feel repetitive or exact-match heavy
- Creating cluster logic that does not match the site’s real navigation
Improve the second draft with implementation feedback
After the first output, return with what you accepted or rejected. For example: “These three pages cannot link due to template constraints,” or “Keep service pages as top-level hubs.” The skill gets noticeably better once the structural constraints are explicit.
Use the templates to standardize team review
If multiple people will act on the output, ask the skill to format every recommendation with the same fields: from, to, anchor, placement, reason, priority. That turns the internal-linking-optimizer guide into an editorial workflow instead of a one-time brainstorm.
Compare proposed links against user journey
Before publishing, check whether a suggested link helps the next logical step for the reader. The best internal links support both SEO and navigation. If a recommendation only exists for keyword adjacency, it is usually a lower-value link.
Re-run after structural changes
Use internal-linking-optimizer again after adding new pillar pages, merging thin content, or changing navigation. Internal linking quality is not a one-time fix; it improves when the skill is used after major content architecture changes.