MCP Setup — AI Agents¶

Connect your AI agent (Claude Code, Cursor, etc.) to your database through SLayer's MCP server. No Python knowledge required.

Prerequisites¶

Install uv — the fast Python package manager. SLayer runs via uvx with no separate install step.

Connect to your agent¶

Claude Code¶

Register SLayer as an MCP server — Claude Code will spawn it automatically when needed:

claude mcp add slayer -- uvx --from 'motley-slayer[embedding_search]' slayer mcp --ingest-on-startup

--ingest-on-startup walks every configured datasource on boot and runs idempotent auto-ingestion before the MCP channel opens, so models are available on the agent's first tool call. Drop it to defer ingestion to a manual ingest_datasource_models call.

The embedding_search extra enables semantic search over models and memories. When it's installed and a provider API key is in the environment (OPENAI_API_KEY by default; override the embedding model with SLAYER_EMBEDDING_MODEL=voyage/voyage-3 + VOYAGE_API_KEY, etc.), the boot-time ingest pass also refreshes per-entity embeddings — hash-skipped, so steady-state boots make zero embedding API calls. Without the extra (or without a provider key), search and ingest still work; the embedding channel is silently disabled.

For databases other than SQLite, add the driver extra alongside (see full list):

claude mcp add slayer -- uvx --from 'motley-slayer[postgres,embedding_search]' slayer mcp --ingest-on-startup

Other agents (JSON config)¶

Most MCP-compatible agents accept a JSON server configuration. Add this to your agent's MCP config file:

{
  "mcpServers": {
    "slayer": {
      "command": "uvx",
      "args": ["--from", "motley-slayer[postgres,embedding_search]", "slayer", "mcp", "--ingest-on-startup"],
      "env": {
        "OPENAI_API_KEY": "sk-..."
      }
    }
  }
}

Replace postgres with your database driver, or use motley-slayer[all] for all supported databases (every driver plus embedding_search).

Remote / shared server¶

SLayer also supports HTTP/SSE transport for running on a different machine, in Docker, or sharing between multiple agents. See the MCP Reference for details.

Verify¶

claude mcp list

Connect a database¶

The recommended approach is to drop a datasource YAML file into your storage folder. This keeps credentials out of the agent conversation and lets you use environment variable references.

Create a file in the datasources/ subdirectory of your storage path (e.g. ~/.local/share/slayer/datasources/mydb.yaml):

name: mydb
type: postgres
host: ${DB_HOST}
port: 5432
database: ${DB_NAME}
username: ${DB_USER}
password: ${DB_PASSWORD}
schema_name: public

${...} references are resolved from environment variables at read time. Set them in your shell before starting the agent, or use a .env file with your agent's environment configuration.

Datasource configs are hot-reloaded — you can add or edit YAML files while the server is running, and the next MCP tool call will pick up the changes. No restart needed.

Once the datasource file is in place, the recommended setup runs MCP with --ingest-on-startup (above), which walks every configured datasource on every boot and runs idempotent auto-ingestion before the MCP channel opens. Models are then available on the agent's first tool call — no ingest_datasource_models call required.

If you didn't start with --ingest-on-startup, fall back to asking your agent:

"Ingest models from the mydb datasource and show me what's available"

The agent will call ingest_datasource_models to generate models from the database schema, then models_summary(datasource_name="mydb") to list them.

You can also create datasources conversationally via the create_datasource MCP tool — see the MCP Reference for details.

Verify it works¶

Ask your agent:

"List the available SLayer models"

The agent should call list_datasources and then models_summary(datasource_name="mydb") and return a list of your tables/models. If it says "no models found", check that:

The --storage path matches where your datasource YAML files are
Models have been ingested (via slayer mcp --ingest-on-startup, ingest_datasource_models, or create_datasource with auto-ingest)
Environment variables referenced in the datasource config are set

Alternative: permanent install¶

If you prefer a traditional install instead of uvx:

uv tool install 'motley-slayer[postgres,embedding_search]'
claude mcp add slayer -- slayer mcp --ingest-on-startup