The Problem with Separate Tools

Most AI coding workflows look like this:

Write notes in Obsidian
Switch to terminal to run Claude
Copy-paste results back into Obsidian
Repeat

Claudian eliminates steps 2 and 3.

A Better Workflow

With Claudian, your Obsidian vault becomes the AI’s working directory. Here’s what that enables:

Research → Implementation in One Step

<!-- In your vault: research-notes.md -->
# API Design Research
- REST vs GraphQL tradeoffs
- Rate limiting strategies
- Authentication patterns

Ask Claudian: “Based on my research notes, generate a TypeScript API client with auth and rate limiting.”

The AI reads your notes and generates code — no copy-pasting.

Plan Mode for Safe Execution

Pressing Shift+Tab enables Plan Mode. The AI will:

Analyze the task
Propose a step-by-step plan
Wait for your approval before executing

This prevents accidental file modifications during exploratory work.

@Mention for Context

Type @ to reference any vault file in your chat:

@architecture-decision-records.md Implement the caching layer we decided on in Q4.

The AI gets the full file content as context.

Slash Commands for Reusable Workflows

Create prompt templates at <vault>/.obsidian/claudian/skills/:

<!-- summarize.md -->
---
description: Summarize and extract action items
---
Summarize the following note and extract action items as a checklist: {{selection}}

Then type /summarize in any conversation.

Example: Daily Note Processing

Here’s a workflow for processing daily notes with AI:

Write rough notes throughout the day in daily/2025-04-28.md
At end of day, open Claudian and type: @daily/2025-04-28.md /summarize
Claudian reads the note, runs the summarize skill, and outputs a clean summary + action items
You accept or refine the output inline

This workflow takes 10 seconds and requires zero context switching.

Getting Started

The Claudian quick-start guide walks through your first conversation in under 5 minutes.