Index research notes with AI
Create or refresh Markdown navigation pages for an existing collection of paper and concept notes while preserving filenames, link targets and local ownership.
You can index research notes with AI when a reading collection has outgrown its original folder list. A concept may appear in several papers, and titles may use more than one language. The generate-mocs skill runs the bundle’s local index builders for paper and concept notes, making the existing collection easier to navigate.
The index is a route into the material you already have. It does not fetch another paper or replace a research note with a new summary. When the collection and output language are unchanged, running it again should leave the generated navigation unchanged as well.
What it does
- Build navigation pages for the configured paper-note folders.
- Create matching indexes for concept categories.
- Preserve actual filenames and existing wikilink destinations.
- Update indexes after adding or reorganising notes.
- Report which files were created, changed or left untouched.
Install and use
- Install the full bundle with Python and its shared index-generation scripts. Configure the local paper and concept directories and grant write access to the chosen index locations.
- Ask, “Refresh the indexes for my reading collection. Keep the existing filenames and tell me which navigation pages changed.”
- Open several links in the updated pages and check that repeating the run on the unchanged collection leaves the results stable.
FAQ
Source
Made by Kiki Huang. Original skill: generate-mocs, published on GitHub under the Apache-2.0 licence.
View on GitHub