Quick Start Tutorial

Quick Start Tutorial

Get SuperLocalMemory working in under 5 minutes — whether you're a new user or upgrading from V2.

---

New Users

1. Install

npm install -g superlocalmemory
# or: pip install superlocalmemory

2. Setup

slm setup

The wizard asks you to pick a mode:

• A (Local Guardian) — Zero cloud. Your data never leaves your machine. Default.
• B (Smart Local) — Local LLM via Ollama for answer synthesis.
• C (Full Power) — Cloud LLM for maximum accuracy. Requires API key.

Most users should start with Mode A — you can switch anytime with slm mode b or slm mode c.

3. Pre-download the embedding model (optional)

slm warmup

Downloads the nomic-embed-text-v1.5 model (~500MB). If you skip this, it downloads automatically on first use.

4. Store your first memory

slm remember "Our API uses JWT tokens with 24-hour expiry. Refresh tokens last 30 days."

Output: Stored 1 facts.

5. Recall it

slm recall "token expiry"

Output shows the stored memory with a relevance score:

  1. [0.82] Our API uses JWT tokens with 24-hour expiry. Refresh tokens last 30 days.

6. Check system status

slm status

SuperLocalMemory V3
  Mode: A
  Provider: none
  Base dir: ~/.superlocalmemory
  Database: ~/.superlocalmemory/memory.db
  DB size: 0.12 MB

7. Check math layer health

slm health

Math Layer Health:
  Total facts: 1
  Fisher-Rao indexed: 1/1
  Langevin positioned: 1/1
  Mode: A

8. Connect to your IDE

slm connect        # Auto-configure all detected IDEs
slm connect --list # See what's configured

Or manually add to your IDE's MCP config:

{
  "mcpServers": {
    "superlocalmemory": {
      "command": "slm",
      "args": ["mcp"]
    }
  }
}

Works with: Claude Code, Cursor, VS Code Copilot, Windsurf, Continue, Cody, ChatGPT Desktop, Gemini CLI, JetBrains, Zed, and more.

9. Open the dashboard

slm dashboard

Opens at http://localhost:8765. 17 tabs: memory browser, knowledge graph, recall lab, trust scores, math health, compliance, and more.

---

Upgrading from V2

If you already have SuperLocalMemory V2 (2.8.6 or earlier):

1. Install V3

npm install -g superlocalmemory

V3 installs alongside V2. Your V2 data is untouched until you migrate.

2. Migrate your data

slm migrate

This will:

• Show your V2 stats (memory count, DB size)
• Ask for confirmation
• Back up your V2 database automatically
• Copy data to the V3 location (~/.superlocalmemory/)
• Convert V2 memories to V3 atomic facts
• Create a symlink so old tools still find the data

3. Setup V3

slm setup     # Choose mode (A/B/C)
slm warmup    # Pre-download embedding model

4. Verify

slm status    # Check V3 is running
slm health    # Check math layers are active
slm recall "something you stored in V2"   # Verify old memories are accessible

What changed from V2 to V3

Feature	V2	V3
Retrieval	Cosine similarity only	4-channel (Semantic + BM25 + Entity Graph + Temporal)
Similarity	Cosine distance	Fisher-Rao geodesic distance
Consistency	None	Sheaf cohomology (algebraic topology)
Lifecycle	Hardcoded thresholds	Self-organizing Langevin dynamics
Modes	Single mode	A (zero-cloud), B (local LLM), C (cloud LLM)
EU AI Act	Not addressed	Mode A/B compliant by design
Dashboard	5 tabs	17 tabs
MCP Tools	6	24
Tests	~200	1400+

Rollback if needed

slm migrate --rollback

This restores your V2 installation. No data is lost.

---

Key Commands Reference

Command	What It Does
slm remember "..."	Store a memory
slm recall "..."	Search memories (semantic + keyword + entity + temporal)
slm forget "..."	Delete matching memories (with confirmation)
slm trace "..."	Recall with per-channel score breakdown
slm status	System status (mode, DB size, path)
slm health	Math layer health (Fisher, Sheaf, Langevin stats)
slm mode a/b/c	Switch operating mode
slm dashboard	Launch web dashboard (http://localhost:8765)
slm mcp	Start MCP server (for IDE integration)
slm connect	Auto-configure IDE integrations
slm profile list	List memory profiles
slm profile create work	Create isolated memory space
slm profile switch work	Switch to a different profile

Full reference: CLI Reference

---

Next Steps

• Modes Explained — Understand A vs B vs C
• MCP Tools — All 24 tools available in your IDE
• IDE Setup — Per-IDE configuration guides
• Auto-Memory — How auto-capture and auto-recall work
• Architecture Overview — How the system works under the hood

---

Part of Qualixar | Created by Varun Pratap Bhardwaj