Preview
What it is.
Ours, and we sell the service it runs inside. A 10-node workflow scrapes your sitemap up to 500 URLs, sorts every URL into priority tiers, then places links into a draft.
Two thresholds do the work, and they are the part people usually guess at: 40 to 50% of links land in the first third of the body, and no paragraph carries more than 2.
On a client with 300+ blog posts the manual version of this audit took a full day. That comparison is the reason to automate it, whoever builds the automation.
What you get.
- A URL categorization pass over the whole sitemap, up to 500 URLs
- Link placement weighted to the first third of the article, at 40 to 50% of total links
- A hard ceiling of 2 links per paragraph
- A repeatable pass across a large archive instead of a one-day manual audit
How to set it up.
Pull the sitemap first and cap the crawl. The version we run stops at 500 URLs, which covers most B2B blogs without a queue.
Categorize every URL before you link anything. Placement quality depends on knowing what each target page is for.
Set the front-load rule at 40 to 50% of links in the first third of the body, so the early paragraphs carry the internal weight.
Cap paragraphs at 2 links. Above that the paragraph reads as a link farm and the anchors stop earning attention.
Give the model the categorized URL list and the draft together, so it picks targets from a known set instead of inventing them.
Spot-check anchors on the first 10 articles. Category-level accuracy is where this pattern fails first.
Pricing Plans
Not sold as a skill.
It runs inside client engagements. The published account dates from April 2026.
Use cases
Linking an unaudited archive
The blog passed a few hundred posts and the manual linking pass stopped happening. Point it at the sitemap; every URL gets sorted into priority tiers before a single link is placed.
Placing links in drafts
A draft is ready and the internal links are guesswork. Hand it the draft together with the categorized URL list; links land weighted toward the first third of the body, never more than a pair per paragraph.
Rules a writer can hold
Writers keep asking where links should go and how many. The front-load share and the per-paragraph cap are two numbers anyone keeps in their head, no tool required.
Handing an engineer the spec
You want your own version built on your own stack. The node count and the crawl ceiling are documented along with both placement thresholds, which is most of a build spec.
Best for
An archive too big to audit by hand
Past a few hundred posts the manual pass stops happening at all, which is the actual problem this pattern solves.
Setting placement rules for a writing team
The front-load percentage and the per-paragraph cap are two numbers a writer can hold in their head without a tool.
Briefing an engineer to build your own version
The node count, the crawl ceiling and both thresholds are documented, which is most of a spec.