Skip to content

User Guide

Complete reference for AI-native project management with Obsidian


Table of Contents

  1. Core Concepts
  2. Entity Types
  3. Workflow Patterns
  4. AI Commands
  5. Plugin Features
  6. Killer Features
  7. Best Practices

Core Concepts

The AI-Human Collaboration Model

This system is designed around a specific workflow:

┌─────────────────────────────────────────────────────────────────┐
│                    Recommended Workflow                         │
├─────────────────────────────────────────────────────────────────┤
│                                                                 │
│  1. Human iterates on design with AI                           │
│     ↓                                                           │
│  2. AI plans the project structure                             │
│     ↓                                                           │
│  3. AI creates specifications (on demand)                       │
│     ↓                                                           │
│  4. Another AI agent implements according to spec              │
│     ↓                                                           │
│  5. Human reviews progress and structure visually              │
│     ↓                                                           │
│  6. Repeat                                                      │
│                                                                 │
└─────────────────────────────────────────────────────────────────┘

Key Principle: - AI does the work (creating, updating, analyzing) - Human reviews visually (canvas, relationships, progress)

Entity Hierarchy

Projects are organized in a 3-tier hierarchy:

Milestone (M-xxx)     ← High-level goals, deliverables
└── Story (S-xxx)     ← Features, user stories
    └── Task (T-xxx)  ← Actionable work items

Plus supporting entities:

Feature (F-xxx)       ← Product features with tier/phase tracking
Decision (DEC-xxx)    ← Architectural decisions
Document (DOC-xxx)    ← Specs, ADRs, guides

Workstreams

All entities belong to a workstream (organizational grouping):

  • engineering - Technical development
  • business - Business logic, product
  • infra - Infrastructure, DevOps
  • research - R&D, experiments
  • design - UX/UI design
  • marketing - Marketing, content

Workstream Normalization: The system automatically normalizes workstream names: - infrastructure, infrainfra - eng, dev, developmentengineering - bizbusiness - ops, operationsoperations - r&d, rndresearch - ux, uidesign - mktgmarketing

Relationships

Entities can have multiple types of relationships:

Relationship Forward Reverse Description
Hierarchy parent children Parent-child structure
Dependency depends_on blocks Task dependencies
Implementation implements implemented_by Story implements Feature/Doc
Documentation documents documented_by Doc describes Feature
Decision Impact - affects Decision affects entities
Supersession supersedes superseded_by Replaces older version

All relationships are bidirectional and auto-synced!

When you set S-001.parent = M-001, the system automatically adds S-001 to M-001.children.


Entity Types

Milestone (M-xxx)

High-level project goals and deliverables.

Key Fields:

id: M-001
type: milestone
title: "Q1 Product Launch"
status: In Progress              # Not Started | In Progress | Completed | Blocked
workstream: engineering
priority: High                   # Low | Medium | High | Critical
target_date: 2026-03-31
owner: john
children: [S-001, S-002]        # Auto-synced

Use Cases: - Quarterly goals - Product releases - Major features - Project phases

Story (S-xxx)

User stories or feature descriptions.

Key Fields:

id: S-001
type: story
title: "User Authentication"
status: In Progress
workstream: engineering
priority: High
parent: M-001                    # Parent milestone
children: [T-001, T-002]        # Auto-synced
depends_on: [S-002]             # Dependencies
acceptance_criteria:
  - Users can sign up with email
  - Users can log in
  - Password reset works

Use Cases: - User stories - Feature implementations - Technical improvements - Bug fixes

Task (T-xxx)

Actionable work items.

Key Fields:

id: T-001
type: task
title: "Implement login endpoint"
status: InProgress              # Open | InProgress | Complete | OnHold
workstream: engineering
parent: S-001                   # Parent story
assignee: john
estimate_hrs: 4
depends_on: [T-002]

Use Cases: - Implementation tasks - Testing tasks - Documentation tasks - Code reviews

Decision (DEC-xxx)

Architectural or design decisions.

Key Fields:

id: DEC-001
type: decision
title: "Use PostgreSQL for primary database"
status: Decided                 # Pending | Decided | Superseded
workstream: engineering
decided_by: john
decided_on: 2026-01-15
affects: [S-001, S-003, DOC-001]  # Entities affected by this decision
context: "Need reliable ACID transactions..."
decision: "Use PostgreSQL 15..."
consequences: "Team needs to learn SQL..."

Use Cases: - Technology choices - Architecture decisions - Design patterns - Process changes

Document (DOC-xxx)

Technical specs, ADRs, guides.

Key Fields:

id: DOC-001
type: document
title: "API Design Specification"
status: Approved                # Draft | Review | Approved | Superseded
workstream: engineering
doc_type: spec                  # spec | adr | guide | reference
version: "1.0"
implemented_by: [S-001, S-002]  # Stories implementing this doc
documents: [F-001]              # Features this documents

Use Cases: - API specifications - Architecture Decision Records (ADRs) - Design documents - Technical guides

Feature (F-xxx)

Product features with tier/phase classification.

Key Fields:

id: F-001
type: feature
title: "Workflow Execution"
status: Complete                # Planned | In Progress | Complete | Deferred
workstream: engineering
tier: OSS                       # OSS | Premium
phase: MVP                      # MVP | 0 | 1 | 2 | 3 | 4 | 5
priority: High
user_story: "As a developer, I want to..."
implemented_by: [M-001, S-001]  # Implementation
documented_by: [DOC-001]        # Documentation

Use Cases: - Product roadmap planning - Feature tracking - Coverage analysis - Release planning


Workflow Patterns

Pattern 1: Design Iteration

Goal: Refine project design with AI before implementation

Human: "I want to build a task management app"
AI:    "Let me help you design it. What are the core features?"
Human: "Users, tasks, projects, and tags"
AI:    "Should we include collaboration features?"
Human: "Yes, but keep it simple for MVP"
AI:    "Here's a proposed structure..."
       [AI suggests milestones and stories]
Human: "Looks good, but split authentication into its own milestone"
AI:    "Updated structure..."

Review: Open canvas to see proposed structure visually

Pattern 2: AI Planning

Goal: Let AI create the full project structure

Human: "Create a project for building a REST API with authentication,
        user management, and data persistence. Target completion in 8 weeks."

AI:    ✓ Created M-001: Authentication System (Week 1-2)
       ✓ Created M-002: User Management (Week 3-4)
       ✓ Created M-003: Data Layer (Week 5-6)
       ✓ Created M-004: API Endpoints (Week 7-8)
       ✓ Created 12 stories across milestones
       ✓ Created 45 tasks with dependencies

Review: Open canvas, run "Populate from vault" and "Reposition nodes"

Pattern 3: Spec-Driven Development

Goal: AI creates specs, another AI implements

Human: "Create a spec for the authentication API"
AI:    ✓ Created DOC-001: Authentication API Specification
       - Endpoints defined
       - Request/response schemas
       - Error handling
       - Security considerations

[Switch to implementation AI agent]

Human: "Implement DOC-001"
AI:    [Reads spec, implements code]
       ✓ Created S-001: Implement auth endpoints (implements DOC-001)
       ✓ Created T-001: POST /auth/signup
       ✓ Created T-002: POST /auth/login
       ✓ Created T-003: POST /auth/refresh
       ✓ All tasks linked to S-001

Review: Check canvas to see implementation structure

Pattern 4: Progress Tracking

Goal: Visual review of project status

Human: "What's the status of M-001?"
AI:    M-001: Authentication System
       Status: In Progress (60% complete)
       - S-001: Complete ✓
       - S-002: In Progress (2/3 tasks done)
       - S-003: Not Started

       Blockers: None

Review: Open canvas to see visual progress, check dependencies


AI Commands

Project Overview

"What's the status of my project?"
"Show me the engineering workstream status"
"What items are blocked?"
"Analyze the project and identify risks"

Creating Entities

"Create a milestone for Q1 launch with target date March 31"
"Create a story under M-001 for user authentication"
"Add a task to S-003 for writing unit tests"
"Create a decision about using PostgreSQL vs MongoDB"
"Create a spec document for the API design"

Updating Entities

"Mark T-005 as completed"
"Update S-002 status to In Progress"
"Assign T-003 to john"
"Set M-001 priority to Critical"
"Archive M-002"

Dependencies

"S-004 depends on S-002 and S-003"
"What's blocking S-006?"
"Show me the dependency graph for M-001"
"Remove the dependency between T-001 and T-002"

Decisions & Documents

"What decisions have been made about authentication?"
"Is DOC-003 up to date with recent decisions?"
"Create an ADR for the database choice"
"Link DEC-001 to S-005"

Features

"What's the feature coverage for Phase 2?"
"Which features are missing documentation?"
"Show me F-001's implementation status"
"Create a feature for user onboarding in tier OSS, phase MVP"

Batch Operations

"Archive all completed stories in M-001"
"Create 5 tasks for S-002: [list of tasks]"
"Update all engineering tasks to assign them to john"

Analysis

"What are the biggest risks in this project?"
"Which workstream has the most blocked items?"
"Show me features without implementation"
"Find circular dependencies"

Plugin Features

Canvas Commands

Access via Command Palette (Ctrl/Cmd+P):

Command Description
Populate from vault Scan vault and add entities to canvas
Reposition nodes Apply hierarchical layout algorithm
Sync edges to dependencies Update frontmatter from canvas edges
Remove duplicate nodes Clean up duplicate entity nodes
Auto-layout features Layout feature entities by tier/phase

Entity Navigator

Navigate relationships via context menu (right-click on entity):

┌─────────────────────────────────┐
│ S-001: User Authentication      │
├─────────────────────────────────┤
│ Go to Parent (M-001)         ▶  │
│ Go to Children               ▶  │
│   └─ T-001: Login endpoint      │
│   └─ T-002: Signup endpoint     │
│ Go to Dependencies           ▶  │
│   └─ S-002: Database setup      │
│ Go to Documents              ▶  │
│   └─ DOC-001: Auth spec         │
│ Go to Decisions              ▶  │
│   └─ DEC-001: JWT tokens        │
└─────────────────────────────────┘

Keyboard Shortcuts: - Ctrl/Cmd+Shift+P - Go to parent - Ctrl/Cmd+Shift+D - Go to documents - Ctrl/Cmd+Shift+E - Go to decisions

Visual Styling

Entities are visually differentiated on canvas:

Entity Border Color Icon
Milestone 3px solid Purple 🎯
Story 2px solid Blue 📖
Task 1px solid Green
Decision 2px dashed Orange ⚖️
Document 1px dotted Yellow 📄
Feature 2px solid Cyan

Workstream Lanes: Entities are organized in horizontal lanes by workstream:

┌─────────────────────────────────────────────────────────┐
│                   Engineering Lane                      │
│  [M-001] ──┬── [S-001] ──┬── [T-001]                   │
│            └── [S-002]   └── [T-002]                    │
├─────────────────────────────────────────────────────────┤
│                   Business Lane                         │
│  [M-002] ──┬── [S-003]                                  │
│            └── [S-004]                                  │
└─────────────────────────────────────────────────────────┘

Archive System

Entities with status: archived are automatically: - Moved to archive/{type}/ folders (e.g., archive/milestones/) - Removed from canvas - Excluded from searches and vault scans

Archive via AI:

"Archive M-001"
"Archive all completed stories"

Restore via AI:

"Restore M-001"


Killer Features

1. Automatic Relationship Syncing

What it does: Maintains bidirectional relationships automatically

Example:

AI: "Set S-001 parent to M-001"

Behind the scenes: - Updates S-001.parent = M-001 - Automatically adds S-001 to M-001.children - Updates both markdown files - Syncs canvas edges

Why it's killer: You never have to manually maintain both sides of a relationship!

2. Transitive Dependency Removal

What it does: Removes redundant dependencies in the graph

Example:

Before:
  C depends on B
  C depends on A
  B depends on A

After (automatic cleanup):
  C depends on B
  B depends on A
  (C→A removed as redundant)

Why it's killer: Keeps dependency graphs clean and understandable!

3. Workstream Normalization

What it does: Automatically normalizes workstream names

Example:

You say: "Create milestone in workstream infrastructure"
System:  ✓ Created with workstream "infra"
         (Normalized: infrastructure → infra)

Why it's killer: Consistent organization without thinking about it!

4. Cycle Detection

What it does: Prevents circular dependencies

Example:

AI: "Make S-001 depend on S-002"
    [S-002 already depends on S-001]

System: ❌ Error: This would create a cycle!
        S-001 → S-002 → S-001

Why it's killer: Catches logical errors before they cause problems!

5. Feature Coverage Tracking

What it does: Tracks which features have implementation and documentation

Example:

AI: "Show feature coverage for Phase MVP"

Response:
  F-001: Workflow Execution
    ✓ Implementation: M-001, S-001, S-002
    ✓ Documentation: DOC-001
    ✓ Tests: 3 files
    Coverage: 100%

  F-002: Tool Invocation
    ✓ Implementation: S-003
    ❌ Documentation: Missing
    ⚠️  Tests: 1 file
    Coverage: 60%

Why it's killer: Know exactly what's done and what's missing!

6. Bulk Operations with Dry-Run

What it does: Preview bulk changes before applying

Example:

AI: "Archive all completed stories in M-001 (dry run)"

Preview:
  Would archive:
  - S-001: User Authentication
  - S-003: Database Setup

  Would update:
  - M-001.children (remove S-001, S-003)

Confirm? (yes/no)

Why it's killer: Safe bulk operations with preview!


Best Practices

1. Use AI for Creation, Plugin for Review

Good:

1. Ask AI to create project structure
2. Open canvas to review visually
3. Ask AI to adjust based on what you see
4. Refresh canvas to verify

Avoid:

Manually creating entities in Obsidian
(Use AI instead - it's faster and maintains relationships)

2. Start with High-Level, Let AI Fill Details

Good:

"Create a milestone for user authentication with stories for
signup, login, and password reset"

Avoid:

"Create milestone M-001"
"Create story S-001"
"Create story S-002"
...
(Let AI create the structure in one go)

3. Use Workstreams Consistently

Good:

All auth-related work: workstream "engineering"
All product decisions: workstream "business"

Avoid:

Mixing workstreams randomly
(Makes visual organization confusing)

4. Document Decisions as You Make Them

Good:

"Create a decision about using JWT for authentication,
affecting S-001 and S-002"

Avoid:

Making decisions without documenting them
(You'll forget why you chose something)

5. Use Dependencies to Show Blockers

Good:

"S-003 depends on S-001 and S-002"
(Now you can ask "What's blocking S-003?")

Avoid:

Not linking dependencies
(Can't track what's blocking what)

6. Archive Completed Work

Good:

"Archive all completed stories in M-001"
(Keeps canvas clean and focused)

Avoid:

Leaving completed work on canvas
(Visual clutter makes it hard to see what's active)

7. Use Features for Product Planning

Good:

Create features for product roadmap
Link stories to features via "implements"
Track coverage with "Show feature coverage"

Avoid:

Planning features without tracking implementation
(Can't see what's done vs planned)


Next Steps


Need Help?