Quick Start

Quick Start Guide

Get up and running with Memory Journal in 2 minutes!

✨ Dynamic Context Management: Once connected, AI agents automatically know when to query your project history and when to create entries - no manual prompting required.

5-Minute Tutorial

Step 1: Install (Choose One)

npm (30 seconds):

npm install -g memory-journal-mcp

npx (zero install):

# Add to MCP config and run directly
npx -y memory-journal-mcp

Docker (1 minute):

docker pull writenotenow/memory-journal-mcp:latest
mkdir data

Step 2: Configure (30 seconds)

Add to ~/.cursor/mcp.json:

npm:

{
  "mcpServers": {
    "memory-journal-mcp": {
      "command": "memory-journal-mcp"
    }
  }
}

Docker:

{
  "mcpServers": {
    "memory-journal-mcp": {
      "command": "docker",
      "args": [
        "run",
        "--rm",
        "-i",
        "-v",
        "./data:/app/data",
        "writenotenow/memory-journal-mcp:latest"
      ]
    }
  }
}

Step 3: Restart Cursor (10 seconds)

The server should show green/connected with:

• 70 tools
• 17 prompts
• 37 resources (24 static + 13 template including dynamic session briefings, health, team resources, GitHub Actions, Milestones, Insights, Kanban, help, gotchas, and flags)

Step 4: Initialize Session Context (NEW in v4.0.0)

For AI clients that don't auto-inject server instructions (or where injection is truncated), read the briefing resource at session start:

Read `memory://briefing/{repo_name}` (or fallback natively to `memory://briefing` if no specific target can be inferred) for project context.

This provides:

• Behavioral guidance (when to create/search entries)
• Latest 3 entries preview
• GitHub status (if configured)
• Quick access links to other resources

Token-efficient: ~300 tokens vs ~6.7K for full instructions.

Step 5: Create Your First Entry (1 minute)

create_entry({
  content:
    "Set up Memory Journal for project context management. Now AI can access complete project history across all threads.",
  entry_type: "technical_achievement",
  significance_type: "milestone",
  tags: ["setup", "context-management", "ai-assistant"],
  is_personal: false,
});

Output:

✅ Created journal entry #1
Type: technical_achievement
Personal: False
Tags: setup, journal, productivity

Step 6: Try Searching (30 seconds)

search_entries({
  query: "journal setup",
  limit: 5,
});

Output:

Found 1 entries:

#1 (milestone) - 2025-10-04 16:30:22
Personal: False
Snippet: Just set up Memory **Journal**! Excited to track my development work...

---

Common First Tasks

Log a Technical Decision (Critical for AI Context)

create_entry({
  content:
    "Decided to use Redis for session caching instead of in-memory store. Rationale: Better scalability for multi-instance deployment. Considered memcached but Redis provides persistence we need for disaster recovery.",
  entry_type: "project_decision",
  tags: ["architecture", "caching", "decision"],
  significance_type: "decision",
});

Build Knowledge Graphs (Connect Related Work)

// Document the implementation
create_entry({
  content:
    "Implemented Redis session caching with 1-hour TTL and automatic refresh on activity.",
  entry_type: "technical_achievement",
  tags: ["caching", "redis", "implementation"],
});
// Returns: Entry #3

// Link implementation to original decision (builds context chain)
link_entries({
  from_entry_id: 3, // Implementation
  to_entry_id: 2, // Original decision
  relationship_type: "implements",
});
// Now AI can trace: Decision → Implementation → Results

Visualize Your Work

// Generate a relationship graph
visualize_relationships({
  entry_id: 2,
  depth: 2,
});

Output: Mermaid diagram showing connected entries!

Backup Your Journal (NEW in v3.0.0)

// Create a named backup
backup_journal({ name: "before_migration" });
// → { success: true, filename: "before_migration.db", sizeBytes: 524288 }

// List available backups
list_backups();
// → { backups: [...], total: 3 }

Prepare Your Daily Standup

Use the prompt palette (type / in Cursor):

/prepare-standup

Output:

📋 Daily Standup - October 4, 2025

What I Did Yesterday:
- Fixed memory leak in caching layer
- Added connection pool monitoring

Blockers/Issues:
(None identified)

What's Next:
- Continue monitoring system performance

Check Server Health (NEW in v3.0.0)

Read the health resource:

memory://health

Returns database stats, backup info, vector index status, and tool filter configuration.

Get Recent Entries

get_recent_entries({
  limit: 5,
  is_personal: false,
});

---

Essential Commands

Creating Entries

// Minimal entry
create_entry_minimal({
  content: "Quick note about the API refactor",
});

// Full entry with all options
create_entry({
  content: "Detailed technical description...",
  entry_type: "project_decision",
  is_personal: false,
  tags: ["api", "refactor", "breaking-change"],
  significance_type: "milestone",
  auto_context: true, // Captures Git info
});

Searching

// Full-text search
search_entries({ query: "API refactor", limit: 10 });

// Semantic search (included by default in v3.0.0)
semantic_search({ query: "performance optimization challenges" });

// Date range
search_by_date_range({
  start_date: "2025-10-01",
  end_date: "2025-10-31",
  tags: ["bug-fix"],
});

Managing Entries

// Update an entry
update_entry({
  entry_id: 1,
  content: "Updated content...",
  tags: ["new-tag"],
});

// Soft delete (recoverable)
delete_entry({
  entry_id: 1,
  permanent: false,
});

// Get specific entry with relationships
get_entry_by_id({
  entry_id: 1,
  include_relationships: true,
});

Analytics

// Get statistics
get_statistics({
  group_by: "week",
});

// List all tags
list_tags();

// Export to Markdown
export_entries({
  format: "markdown",
  start_date: "2025-10-01",
  tags: ["milestone"],
});

---

Using Workflow Prompts

Access prompts through your MCP client's prompt palette (type / in Cursor):

Daily Standup

/prepare-standup

Sprint Retrospective

/prepare-retro days:14

Weekly Digest

/weekly-digest

Period Analysis

/analyze-period start_date:2025-10-01 end_date:2025-10-31

Goal Tracking

/goal-tracker

Find Related Entries

/find-related query:"context management"

---

Tips for Success

Entry Types for AI Context Management

• project_decision - Technical decisions, architecture choices (MOST IMPORTANT for AI context)
• technical_achievement - Implementations, breakthroughs
• feature_implementation - Feature builds, rollouts, improvements
• bug_fix - Bug resolutions with root cause analysis
• research - Research, exploration, investigation
• learning - Learning insights, patterns, documentation
• planning - Planning sessions, roadmaps, specs
• retrospective - Sprint and project retrospectives
• personal_reflection - Personal thoughts and notes (optional)

Tagging Strategy

Be consistent:

• Use lowercase
• Use hyphens for multi-word tags
• Create a core set of tags you use regularly

Good tags:

• performance, bug-fix, api-design
• learning, documentation, testing
• Project-specific: memory-journal, postgres-mcp

Relationship Types

• references - General connection between entries
• implements - Implementation of a spec/plan
• clarifies - Explanation or elaboration
• evolves_from - Iteration or improvement
• response_to - Reply or follow-up
• blocked_by - Entry was blocked by another (causal)
• resolved - Entry resolved/fixed another (causal)
• caused - Entry caused/led to another (causal)

Auto-Context Capture (Critical for AI Context)

Set auto_context: true (default) to automatically capture Git/GitHub state with every entry:

• Repository context - Repo name, path, working directory
• Branch information - Current branch (e.g., feature/redis-caching)
• Commit details - Latest commit hash and message
• GitHub integration - Open issues, PR status (if gh CLI installed)

Why this matters: AI can understand WHEN and WHERE work happened, making context reconstruction much more accurate across fragmented threads.

---

Next Steps

Learn More:

• Tools Reference - Complete guide to all 70 tools
• Tool Filtering - Reduce token usage by up to 84%
• Prompts Guide - All 17 workflow prompts
• Visualization - Relationship graphs
• Examples - Real-world usage patterns

Advanced Features:

• Relationships - Building knowledge graphs
• Semantic Search - ML-powered search
• Analytics - Statistics and insights

---

Questions? Check the Examples page or open an issue on GitHub.