MCP tools and workflows
Explore six current SQL Cockpit MCP Server tools and related table-synchronisation workflow guidance.
Available tools
list_sync_configs
List SQL Tables Sync configuration rows, optionally filtering to enabled entries.
Inputs: enabledOnly (optional boolean)
REST call: GET /api/configs
get_sync_config
Retrieve one configuration by ID or match a sync name from the configuration list.
Inputs: syncId (integer) or syncName (string)
REST call: GET /api/configs/:id or GET /api/configs
get_table_schema
Find a selected table and its columns in database metadata.
Inputs: connection (object), schema (string), table (string)
REST call: POST /api/databases/metadata
get_table_batch_size_recommendation
Request advisory batch-size guidance for a table.
Inputs: connection (object), schema (string), table (string)
REST call: POST /api/tables/batch-size-recommendation
generate_table_migration
Generate destination migration SQL by comparing supplied source and destination tables; does not execute it.
Inputs: source (object), destination (object), options (optional object)
REST call: POST /api/migrations/table-diff
generate_sync_migration
Generate destination migration SQL from a sync configuration; does not execute it.
Inputs: syncId (integer) or syncName (string); includeDropColumns, includeAlterColumns, includePrimaryKey (optional booleans)
REST call: POST /api/migrations/from-config
The current tools do not execute SQL, start synchronisations, monitor jobs or validate outcomes.
Workflow explainers
These broader operational workflows are not additional callable tools.
- Inspect table structure - Establish which schemas, tables, columns and data types are in scope before comparing or moving data.
- Compare environments - Identify schema drift, missed table changes and dependencies that could affect a synchronisation or migration window.
- Prepare a synchronisation - Turn a proposed table move into a reviewable scope with an approval decision and validation criteria.
- Request approved execution - Describe the guardrails a data-changing request needs before Engine performs a synchronisation.
- Check synchronisation progress - Connect current operation status with job history, retries and movement evidence while work is in progress.
- Validate the outcome - Collect post-run comparison and sign-off evidence instead of treating a completed job as proof that data is correct.
- Investigate runtime changes - Explain a slow, failed or intermittent synchronisation using job history and operational baselines.