Building Related Recommendations with LLM-WIKI Skill

Preface

I often see some blogs that have a "Related Recommendations" section at the end of each post.

A few days ago I tried out the LLM-WIKI skill, so I thought about using the LLM-WIKI skill to analyze my blog and create related recommendations for each post.

Approach

Save my blog pages as WIKI pages,

Use the LLM-WIKI skill to lint and analyze these pages, establishing links between them,

Count the link counts of all pages,

For each page, among the pages it links to, find the top 3 pages with the highest link counts.

Practice

Initialize the wiki

I had a chat with hermes (MIMO v2 pro 2026-4-21)

I plan to do this task, help me analyze the specific plan, don't implement it yet, let's discuss first

There is a website

https://zelikk.blogspot.com/

This website's RSS output contains full content

https://zelikk.blogspot.com/rss.xml

Use your llm-wiki skill to create a new WIKI

Ingest every article from this website into this WIKI

Lint and organize this WIKI,

and output the following results

1. For each link in the WIKI, which page links to which page

2. The ranking of inbound link counts for each page

inbound link - a page being linked to by other pages

3. Among the outbound-linked pages of each page, the top 3 pages with the highest counts in list 2, and the respective counts of these 3 pages

outbound link - a page linking to other pages

The entire development and communication process involved a lot of detailed design, so I won't list them all. See github for details

In the end, I got these files

1. Page assignment index

index,page,pub_date,raw_url

1,hermes-python,2026-04-21,https://zelikk.blogspot.com/2026/04/hermes-python.html

2. List of links between pages

source,target

1,4

3. Link relationships of a single page

rank,index,page,degree,neighbors

1,98,xray-vless-reality-tls-x25519,47,"29|43|74|102|103|131|132|133|144|147|149|150|153|154|155|157|158|160|162|173|174|181|191|194|207|225|243|244|266|268|272|284|287|288|300|308|310|315|321|323|324|338|339|340|422|504|509"

4. The top 3 pages with the highest link counts among the pages linked by a page

index,page,top1_index,top1_page,top2_index,top2_page,top3_index,top3_page

1,hermes-python,13,hermes-agent-oracle-vps-ubuntu-root,10,hermes-agent,4,hermes-agent-telegram-group-topic-agent

The semantic analysis of all my blog articles took quite a long time and used a lot of tokens. Fortunately, hermes could freeload Xiaomi MIMO v2 pro at the time.

While hermes was busy, I used the free web version of Claude to design a single-page HTML tool.

(Upload the 1. page assignment index file page_index.csv, upload the 4. file link_outbound_top3.csv with the top 3 pages with the highest link counts) and said:

I have these 2 data source files

I need a single HTML page file that displays the content of link_outbound_top3.csv in a different way

Take

1,hermes-python,13,hermes-agent-oracle-vps-ubuntu-root,4,hermes-agent-telegram-group-topic-agent,10,hermes-agent

as an example,

I want to look up the content of raw_url in page_index.csv based on the index, and get this result

https://zelikk.blogspot.com/2026/04/hermes-python.html

https://zelikk.blogspot.com/2026/04/hermes-agent-oracle-vps-ubuntu-root.html

https://zelikk.blogspot.com/2026/04/hermes-agent-telegram-group-topic-agent.html

https://zelikk.blogspot.com/2026/04/hermes-agent.html

This display tool has also been uploaded to github

With the help of this tool, I manually edited my blog post by post, adding things like "Related Recommendations", "Featured Recommendations", "Want to See More", "See Other Stuff" at the end of each post...

Actually, friends who understand code, after reading this far, certainly know that there are countless ways to achieve automatic display from such data. It's just that my interest isn't there, so I didn't continue to put in more effort.

Maintaining the wiki

Two days later, my blog had 2 new posts, so there were 2 new entries in the RSS.

I asked hermes to summarize the previous process into a skill, then use it to process the new RSS of my blog, and it indeed errored out.

Then there was continuous dialogue to fix the issues.

btw, hermes's freeload Xiaomi model event has ended, and there's a new freeload moonshotai/kimi-k2.6

It can be considered complete in the end. I learned this skill under a new hermes system, then applied it to another person's blog, and it worked normally.

========