learner `v1.1.0`

Capture and organise personal study notes by topic in structured Markdown files.

Overview

The learner plugin provides a conversational agent that helps you build and maintain a personal knowledge base. For every topic you study, the agent creates or updates a dedicated Markdown file with a consistent header, a table of contents placeholder, and individually titled sections separated by horizontal rules. Sections can include typed text, pasted content, attached images, and auto-generated Mermaid diagrams.

Each session follows a guided flow:

Identify the topic (or provide it in your opening message)
Optionally specify a sub-directory under your notes folder (any depth, e.g. work/docker)
Choose to update an existing note or start a fresh one
Add or continue named sections at your own pace
Confirm when each section is done — the agent writes and saves it immediately
Optionally enrich sections with Mermaid diagrams (suggested automatically when relevant)

Prerequisites

No additional installation is required. The plugin works with the VS Code Chat Plugin Marketplace out of the box.

To customise the notes directory, set the LEARNER_NOTES_DIR environment variable before starting VS Code:

# Linux / macOS
export LEARNER_NOTES_DIR="$HOME/work/learner-notes"

# Windows (PowerShell)
$env:LEARNER_NOTES_DIR = "C:\work\learner-notes"

If LEARNER_NOTES_DIR is not set, the agent falls back to:

Platform	Default path
Windows	`%USERPROFILE%\Documents\LearnerNotes`
Linux / macOS	`$HOME/Documents/LearnerNotes`

Hooks

Event	Script	What it does
`SessionStart`	`inject-learner-notes-dir.ps1` / `.sh`	Reads `LEARNER_NOTES_DIR` (or uses the platform default) and injects the resolved path into the agent context as `additionalContext`.

Agents

Agent	File	Description
Topic Scriber	`TopicScriber.agent.md`	Interactive agent that guides the learner through creating and updating topic-based study notes, capturing text, images, and Mermaid diagrams in structured Markdown files.

Output Document Structure

Each topic note follows this structure:

# <Topic>

|Learner|Date|
|--------|------|
|<learner_name>|<current_date>|

-----

[TOC]

-----

## <Section Title>

<Section content>
▸ Mermaid diagram (when applicable)

-----

## <Another Section Title>

<Section content>

-----

File Naming

Topic files are always lowercase, with spaces replaced by hyphens, and stripped of non-alphanumeric/hyphen characters:

Topic Name	Filename
`Machine Learning Basics`	`machine-learning-basics.md`
`Python Basics!`	`python-basics.md`
`Docker Networking`	`docker-networking.md`

Configuration Reference

Environment Variable	Default	Description
`LEARNER_NOTES_DIR`	`%USERPROFILE%\Documents\LearnerNotes` (Windows) / `$HOME/Documents/LearnerNotes` (Linux/macOS)	Root directory where all topic note files are stored.

Sub-directory Support

At the start of each session the agent asks whether you want to save notes in a sub-directory under the resolved notes folder. The sub-directory can be any number of levels deep (e.g. work/docker or school/math/calculus). If you leave it blank, notes are saved directly in the root notes folder. The agent creates all intermediate directories automatically.

Author

Dimpletz

learner v1.1.0