Getting Started

1

Get your API key

Sign up at smara.io to get your free API key. 10,000 memories included.

Create a free account and grab your API key from the dashboard. Your key looks like this:

Format: smr_xxxxxxxxxxxx

Keep this key handy — you'll paste it into your config in the next step.

Sign up free on smara.io Takes 30 seconds. No credit card required.

2

Install the MCP server

Add Smara to your AI tool in 30 seconds.

Add this to your claude_desktop_config.json:

claude_desktop_config.json

{
  "mcpServers": {
    "smara": {
      "command": "npx",
      "args": ["@smara/mcp-server"],
      "env": {
        "SMARA_API_KEY": "smr_YOUR_KEY_HERE"
      }
    }
  }
}

Add this to .cursor/mcp.json in your project root:

.cursor/mcp.json

{
  "mcpServers": {
    "smara": {
      "command": "npx",
      "args": ["@smara/mcp-server"],
      "env": {
        "SMARA_API_KEY": "smr_YOUR_KEY_HERE"
      }
    }
  }
}

Add this to your Windsurf MCP config:

~/.windsurf/mcp.json

{
  "mcpServers": {
    "smara": {
      "command": "npx",
      "args": ["@smara/mcp-server"],
      "env": {
        "SMARA_API_KEY": "smr_YOUR_KEY_HERE"
      }
    }
  }
}

Run the auto-installer which detects your tools and configures everything:

Terminal

npx @smara/mcp-server --init

This detects installed AI tools and writes the config automatically. You'll be prompted for your API key.

3

Store your first memory

Just start coding. Smara auto-captures context.

Once installed, your AI tool stores memories automatically. Here's what it looks like:

Y

Let's use PostgreSQL for the user service. We need JSONB for the preferences column and a GIN index on tags.

AI

Got it. I'll set up PostgreSQL with JSONB for preferences and add a GIN index on the tags column.

Stored to Smara

Or store manually via the API:

Terminal

curl -X POST https://api.smara.io/v1/memories \
  -H "Authorization: Bearer smr_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "content": "We chose PostgreSQL with JSONB for preferences",
    "source": "architecture-decision",
    "importance": 8
  }'

4

Search your memories

Ask your AI about past decisions.

Your AI automatically searches Smara when you ask about past context:

Y

What did we decide about the database schema?

AI

Based on our earlier discussion, you chose PostgreSQL with a JSONB column for user preferences and a GIN index on the tags column.

Found via Smara

Or search directly via the API:

Terminal

curl https://api.smara.io/v1/memories/search \
  -H "Authorization: Bearer smr_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query": "database schema decisions"}'

5

Set up a team optional

Share project context with your team.

Teams let you share project decisions with collaborators while keeping personal preferences private. Your AI automatically classifies memories as team-shared or personal.

Architectural decisions, API patterns → shared with team
Editor preferences, personal workflows → stay private
Combined search returns both when relevant

Teams documentation Learn about roles, permissions, and team memory management.

6

Explore the playground

Try the API live in your browser.

The playground lets you test every API endpoint directly in your browser. Store memories, search, and experiment with decay scoring — all without writing code.

Open the Playground Interactive API explorer with live requests.

Get started with Smara

All done! You're set up.