ArcadeDB Multi-Model DBMS
OfficialServer Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| ARCADEDB_ROOT_PASSWORD | Yes | Root user password for the ArcadeDB server (used for MCP authentication). |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| list_databasesA | List all databases available to the authenticated user on the ArcadeDB server. |
| get_schemaB | Get the full schema of a database including types (vertex, edge, document), their properties, indexes, and inheritance hierarchy. |
| queryA | Execute a read-only (idempotent) query against an ArcadeDB database. Use this for SELECT, MATCH, and other read operations. Prefer OpenCypher (language: 'cypher') unless SQL is explicitly requested. |
| execute_commandA | Execute a non-idempotent command against an ArcadeDB database. Use this for INSERT, UPDATE, DELETE, CREATE TYPE, and other write operations. Prefer OpenCypher (language: 'cypher') unless SQL is explicitly requested. |
| server_statusA | Get ArcadeDB server information including version, server name, available query languages, and HA/cluster status. |
| profiler_startA | Start the query profiler to capture query execution data. The profiler records all queries with their execution times and plans. It auto-stops after the specified timeout (default 60 seconds). Use profiler_stop to stop early and get results, or profiler_status to check progress. |
| profiler_stopA | Stop the query profiler and return the captured profiling data. Returns aggregated query statistics including execution counts, timing (min/max/avg/p99), and execution plan step costs. Results are also saved to disk for later retrieval. |
| profiler_statusA | Get the current status of the query profiler. Returns whether the profiler is recording, and if available, the current or last profiling results including captured queries, timing statistics, and server metric snapshots. |
| get_server_settingsA | Retrieve ArcadeDB server configuration settings. Returns all server-level settings with their current values, defaults, and descriptions. Sensitive values (passwords) are masked. Use this to understand and diagnose server configuration. |
| set_server_settingA | Update a server configuration setting at runtime. Changes take effect immediately but may not persist across server restarts (depends on the setting). Use get_server_settings first to see available settings and their current values. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 10 tools
Each tool targets a clearly distinct concern: server admin (status, settings), profiler lifecycle (start/stop/status), and database operations (list, schema, query, execute). No two tools overlap in purpose, making misselection unlikely.
Tools follow a consistent verb_noun pattern: profiler_start/stop/status, get_server_settings, set_server_setting, list_databases, get_schema, execute_command, query, server_status. The only slight deviation is 'execute_command' and 'query' which use more generic verbs rather than a resource-specific noun, and 'server_status' uses a different word order than get_/set_ prefixes.
Ten tools is well-scoped for a DBMS server MCP. Each tool maps to a meaningful operation a client would need: server introspection, config management, profiling, and database CRUD/querying. Nothing feels redundant or missing at the top level.
The surface covers server status, configuration management, profiling, database enumeration, schema retrieval, and both read and write operations. Minor gaps include no database creation/deletion and no explicit way to retrieve saved profiler results from disk, but the core admin+query workflow is complete.