Skip to main content
Glama
timescale

Tiger Memory MCP Server

Official
by timescale
README.md
# Tiger Memory MCP Server

A simple memory system designed to allow LLMs to store and retrieve information. This provides some focused tools to LLMs via the [Model Context Protocol](https://modelcontextprotocol.io/introduction).

## API

All methods are exposed as MCP tools and REST API endpoints.

## Development

Cloning and running the server locally.

```bash
git clone git@github.com:timescale/tiger-memory-mcp-server.git
```

### Building

Run `npm i` to install dependencies and build the project. Use `npm run watch` to rebuild on changes.

Create a `.env` file based on the `.env.sample` file.

```bash
cp .env.sample .env
```

### Testing

The MCP Inspector is very handy.

```bash
npm run inspector
```

| Field          | Value           |
| -------------- | --------------- |
| Transport Type | `STDIO`         |
| Command        | `node`          |
| Arguments      | `dist/index.js` |

#### Testing in Claude Desktop

Create/edit the file `~/Library/Application Support/Claude/claude_desktop_config.json` to add an entry like the following, making sure to use the absolute path to your local `tiger-memory-mcp-server` project, and real database credentials.

```json
{
  "mcpServers": {
    "tiger-memory": {
      "command": "node",
      "args": [
        "/absolute/path/to/tiger-memory-mcp-server/dist/index.js",
        "stdio"
      ],
      "env": {
        "PGHOST": "x.y.tsdb.cloud.timescale.com",
        "PGDATABASE": "tsdb",
        "PGPORT": "32467",
        "PGUSER": "readonly_mcp_user",
        "PGPASSWORD": "abc123"
      }
    }
  }
}
```

## Deployment

We use a Helm chart to deploy to Kubernetes. See the `chart/` directory for details.

The service is accessible to other services in the cluster via the DNS name `tiger-memory-mcp-server.savannah-system.svc.cluster.local`.

### Database setup

Creating the database user:

```sql
CREATE USER tiger_memory WITH PASSWORD 'secret';
GRANT CREATE ON DATABASE tsdb TO tiger_memory;
```

### Secrets

Run the following to create the necessary sealed secrets. Be sure to fill in the correct values.

```bash
kubectl -n savannah-system create secret generic tiger-memory-mcp-server-database \
  --dry-run=client \
  --from-literal=user="tiger_memory" \
  --from-literal=password="secret" \
  --from-literal=database="tsdb" \
  --from-literal=host="x.y.tsdb.cloud.timescale.com" \
  --from-literal=port="32467" \
  -o yaml | kubeseal -o yaml

# https://logfire-us.pydantic.dev/tigerdata/tigerdata/settings/write-tokens
kubectl -n savannah-system create secret generic tiger-memory-mcp-server-logfire \
  --dry-run=client \
  --from-literal=token="pylf_v1_us_" \
  -o yaml | kubeseal -o yaml

# https://login.tailscale.com/admin/settings/keys
kubectl -n savannah-system create secret generic tiger-memory-mcp-server-tailscale \
  --dry-run=client \
  --from-literal=authkey="tskey-auth-" \
  -o yaml | kubeseal -o yaml
```

Update `./chart/values/dev.yaml` with the output.

TDQS

A3.7/5.0

Scored across 4 tools

Disambiguation5/5

Each tool has a unique, non-overlapping purpose: create (remember), read (recall), update (update), delete (forget). No ambiguity.

Naming Consistency5/5

All tool names are single-word imperative verbs (forget, recall, remember, update), following a consistent pattern.

Tool Count5/5

4 tools cover the essential CRUD operations for memory management, which is well-scoped and appropriate.

Completeness5/5

The tool set fully covers all CRUD operations (Create, Read, Update, Delete). No obvious gaps for the stated purpose.

Maintenance

ActivityStale
ResponsivenessUnresponsive