# Diceroll — Atomic Note Discovery
Pick a random atomic note from `2 Resources/Topics/` for serendipitous rediscovery.
## Usage
```
/diceroll
```
## Instructions
### Step 1: Select a random atomic note
Run: `find "2 Resources/Topics" -name "*.md" -not -name "README.md" -type f | shuf -n1`
If the pool is empty, show: "No atomic notes found in Topics."
### Step 2: Extract context
From the selected file path (`2 Resources/Topics/[Topic Name]/[Note].md`):
- **Note title**: The H1 heading from the file, or filename without extension if no H1
- **Topic name**: The parent directory name under `2 Resources/Topics/`
- **Cross-references**: Count and collect all `[[wikilinks]]` in the note
### Step 3: Display the note
Show the note in this exact format:
```
## [Note Title]
([Topic Name] | X cross-references)
[Core insight — first 1-2 substantive paragraphs from the note, after the H1/Overview heading. Skip metadata blocks, empty lines, and heading lines to find the actual prose content.]
Cross-references: [[Note A]], [[Note B]], [[Note C]]
```
Rules:
- No metadata dump (no PARA category, size, creation date)
- No numbered menu
- Show the note's actual content, not a summary about it
- If there are no cross-references, omit the Cross-references line
- Keep the core insight to 1-2 paragraphs max — enough to re-engage with the idea