Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@starrocks-mcp Show me the schema for the campaign_budget table"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
A read-only MCP (Model Context Protocol) server for StarRocks databases. Query and explore your StarRocks data through Claude and other AI assistants.
SQL Query Execution
Execute SQL queries directly from Claude Desktop
Automatic result truncation for large datasets
Table Operations
List all databases
List tables in a database
Get table schema details
LDAP Authentication
Secure LDAP authentication support
Each user uses their own credentials
Data Ingestion
Table/Schema Creation or Modification
User/Permission Management
Related MCP server: imply-druid-mcp
Tool | Description |
| Execute a SQL query and return results |
| List all databases in StarRocks |
| List tables in a database |
| Get detailed schema information for a table |
uvx --from git+https://github.com/jason-ung/starrocks-mcp starrocks-mcppip install git+https://github.com/jason-ung/starrocks-mcp
starrocks-mcpEnvironment Variable | Required | Default | Description |
| Yes | - | StarRocks host |
| No |
| StarRocks port |
| Yes | - | Your LDAP username |
| Yes | - | Your LDAP password |
| No |
| Default database |
| No |
| Logging level |
| No |
| Default query timeout (ms) |
| No |
| Maximum SQL query length |
Add to your Claude Desktop config file:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"starrocks": {
"command": "uvx",
"args": [
"--from",
"git+https://github.com/jason-ung/starrocks-mcp",
"starrocks-mcp"
],
"env": {
"STARROCKS_HOST": "starrocks.datapl.datahou.se",
"STARROCKS_PORT": "9030",
"STARROCKS_USER": "your.ldap.username",
"STARROCKS_PASSWORD": "your_ldap_password",
"STARROCKS_DATABASE": "ads"
}
}
}
}⚠️ Important: Replace your.ldap.username and your_ldap_password with your actual LDAP credentials.
User: Show me all databases in StarRocks
Claude: [Uses list_databases]
- ads
- search
- user_behaviorUser: What tables are in the ads database?
Claude: [Uses list_tables with database="ads"]
- advtr_product_da_preprocessed_log
- advtr_product_sa_preprocessed_log
- campaign_budgetUser: Describe the advtr_product_da_preprocessed_log table
Claude: [Uses describe_table with table="advtr_product_da_preprocessed_log"]
Field | Type | Description
----- | ---- | -----------
eventTs | BIGINT | Event timestamp
eventType | VARCHAR | Event type (IMPRESSION, CLICK, etc.)
...User: Query the top 10 campaigns by impression count
Claude: [Uses execute_query]
SELECT campaignId, COUNT(*) as impression_count
FROM advtr_product_da_preprocessed_log
WHERE eventType = 'IMPRESSION'
AND base_dt = '2026-01-29'
GROUP BY campaignId
ORDER BY impression_count DESC
LIMIT 10git clone https://github.com/jason-ung/starrocks-mcp.git
cd starrocks-mcp
# Create virtual environment
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
# Install dependencies
pip install -e ".[dev]"
# Copy .env.example to .env and configure
cp .env.example .env
# Edit .env with your credentials
# Run the server
starrocks-mcp# Test StarRocks connection
mysql -h starrocks.datapl.datahou.se -P 9030 -u your.username -p --ssl-mode=DISABLED --enable-cleartext-pluginNever commit credentials: Always use environment variables for sensitive data
Use personal LDAP accounts: Each user should use their own credentials
LDAP authentication: Supports cleartext LDAP authentication over secure connections
Read-only operations: This MCP server only supports SELECT queries and metadata operations
MIT License - see LICENSE for details.
Jason Son (@jason-ung)
Made with ❤️ for Bucketplace Ads Team
Get table schema and column information
| Name | Required | Description | Default |
|---|---|---|---|
| table | Yes | Table name | |
| database | No | Database name (optional, defaults to 'ads') |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure, yet it only states the basic action. It does not clarify that this is a safe read-only operation, what format the schema information returns in, or how errors (e.g., table not found) are handled.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is extremely concise at five words, with the action verb front-loaded. There is no redundancy or wasteful text—every word contributes to understanding the tool's purpose.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple two-parameter introspection tool with complete schema coverage, the description is minimally adequate. However, given the lack of annotations and output schema, it could have benefited from mentioning the specific information returned (data types, constraints, indexes).
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, meaning the schema already documents both 'table' and 'database' parameters including the default value. The description adds no additional parameter context, but the baseline score of 3 is appropriate given the complete schema coverage.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states a clear action ('Get') and resource ('table schema and column information'), accurately describing the tool's function. However, it lacks explicit differentiation from sibling tools like 'list_tables', which could confuse agents about whether this returns metadata or just names.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides no guidance on when to use this tool versus alternatives (e.g., when to use 'describe_table' vs 'execute_query' with a DESCRIBE statement). There are no prerequisites, workflow hints, or explicit exclusions mentioned.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Execute a SQL query against StarRocks and return results.
| Name | Required | Description | Default |
|---|---|---|---|
| sql | Yes | SQL query to execute |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations provided, so description carries full burden, yet it fails to disclose critical behavioral traits: whether the tool supports write operations (INSERT/UPDATE/DELETE), read-only safety, result format/size limits, or execution timeouts. 'Return results' is vague given the high-stakes nature of arbitrary SQL execution.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Single sentence, front-loaded with action verb, no redundancy. However, given the tool's high complexity and lack of annotations/output schema, the brevity crosses from concise to under-specified.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Arbitrary SQL execution is high-complexity with significant safety implications. With no annotations, no output schema, and minimal description, the definition lacks essential context about data modification risks, result structure, and operational scope required for safe agent invocation.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema has 100% coverage with clear parameter description. The description adds context by specifying 'StarRocks' as the target system (not in schema), but adds no detail on SQL syntax requirements, supported dialect features, or parameter binding beyond the schema's basic declaration.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Clear specific verb ('Execute') and resource ('SQL query') against StarRocks. However, it does not distinguish from sibling metadata tools (describe_table, list_tables), which also retrieve data but via specific patterns rather than arbitrary SQL.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Provides no guidance on when to use this versus the specific metadata tools (list_tables, describe_table) or when to prefer built-in tools over raw SQL. No mention of prerequisites or query complexity considerations.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List all databases in StarRocks
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full disclosure burden. While 'List' implies read-only behavior and 'all' defines scope, it omits details about return format, pagination limits, or required permissions to view all databases.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Five words, zero redundancy. Front-loaded with action verb ('List') followed immediately by target resource. Every word earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given zero parameters and no annotations, the description adequately covers the tool's function. Minor gap: without an output schema, it could briefly mention what data is returned (database names vs metadata), but the scope is sufficiently clear for a simple listing operation.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Input schema has zero parameters. Per rubric rules, zero parameters establishes a baseline score of 4. The description correctly implies no filtering parameters are accepted.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description provides a specific verb ('List'), clear resource ('databases'), and scope ('all...in StarRocks'). It clearly distinguishes from sibling 'list_tables' (databases vs tables) and 'describe_table'/'execute_query' (listing vs describing/executing).
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description states what the tool does but provides no explicit guidance on when to use this versus siblings like 'list_tables' (e.g., 'use this first to find database names before listing tables') or any prerequisites needed to access database metadata.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List tables in a database
| Name | Required | Description | Default |
|---|---|---|---|
| database | No | Database name (optional, defaults to 'ads') |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden of behavioral disclosure. While 'List' implies read-only behavior, the description does not confirm safety, disclose return format (e.g., array of table names vs objects), mention pagination limits, or note the default 'ads' database behavior.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is extremely terse at only 5 words, but it efficiently conveys the core operation without redundancy. However, it is so minimal that it fails to front-load any distinguishing characteristics or behavioral constraints that would help tool selection.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a tool with no output schema and no annotations, the description should disclose what the user receives (e.g., 'returns table names') and highlight the default database behavior. As written, it leaves significant gaps in the agent's understanding of the tool's contract.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Input schema has 100% description coverage, establishing a baseline of 3. The tool description adds no semantic context about the 'database' parameter (e.g., not mentioning that it defaults to 'ads'), but no additional documentation is required given the complete schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the verb (List), resource (tables), and scope (in a database). However, it does not explicitly differentiate from sibling tools like 'describe_table' or 'list_databases', which would help the agent choose between listing metadata versus detailed schema information.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides no guidance on when to use this tool versus alternatives like 'describe_table' (which gets details for one table) or 'list_databases' (which scopes higher). No prerequisites or conditions are mentioned.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.1.0describe_tableexecute_querylist_databaseslist_tablesScored across 4 tools
Each tool has a clearly distinct purpose: describe_table focuses on schema details, execute_query handles SQL execution, list_databases enumerates databases, and list_tables lists tables within a database. There is no overlap or ambiguity between these functions.
All tool names follow a consistent verb_noun pattern (describe_table, execute_query, list_databases, list_tables) with clear, descriptive verbs and nouns. There are no deviations in naming style.
With 4 tools, the count is reasonable for a database interaction server, covering core operations like listing, describing, and querying. It might be slightly thin for advanced database management, but it's well-scoped for basic functionality.
The tools cover essential read and query operations (list, describe, execute), but there are notable gaps for a database server, such as create/update/delete operations for databases or tables, which could limit agent workflows for full lifecycle management.
Query your org's data in natural language — read-only MCP access to SQL, NoSQL, files & warehouses.
Draxlr's remote MCP server connects AI assistants to your SQL databases and dashboards. Explore schemas, run read-only queries, manage saved queries and dashboards, and export results, all with row-level security so each user sees only their own data.
Query your warehouse or a CSV with Claude/ChatGPT over MCP, governed by table-level ACL + audit.
Cloud-hosted MCP server for secure AI access to enterprise data sources via CData Connect AI.
