ahastudio/geeknews-hot

Pick popular GeekNews articles and, for each one, write a TIL document in the same format as the analyze-article skill.

Usage

/geeknews-hot
/geeknews-hot 3

• Argument: number of documents to write (default: 5)

Procedure

1. Collect popular articles

Fetch https://news.hada.io/ with WebFetch. Extract the list of articles ordered by points, highest first. Collect each article's title, source URL, point count, and GeekNews topic URL.

2. Remove duplicates

Use Glob and Grep to search existing TIL documents and exclude any article that already has a document on the same topic. Judge duplicates by title, source URL, project name, etc.

After excluding duplicates, pick the requested number of articles by point order.

3. Report the selection

Show the selected list of articles to the user. Format:

1. Title (points) — expected path
2. Title (points) — expected path
...

If any articles were excluded, show them along with the reason.

4. Gather materials

For each selected article, in parallel:

• Fetch the source URL content with WebFetch.
• Fetch the body summary and comments of the GeekNews topic page with WebFetch.

5. Write the documents

For each selected article, invoke the /analyze-article skill with the source URL and the chosen file path as arguments.

Additional context to pass when invoking:

• Include the fetched GeekNews comments so they can be reflected in the analysis or critique.
• Never recreate a document that already exists.

Create the file in a directory that matches the topic. Criteria for choosing the directory:

• Refer to the existing TIL directory structure.
• Place it in the most appropriate existing directory.
• If there is no suitable existing directory, do NOT create a new one — ask the user.