MCP Tools

AWS MCP Server (recommended)

When connected, the AWS MCP Server provides:

Knowledge tools (no extra setup beyond the server itself):

• aws___search_documentation — search across all AWS documentation, including DSQL service docs and skills. PREFER for verifying DSQL limits or finding the canonical doc page.
• aws___read_documentation — fetch a specific AWS docs page in markdown form.
• aws___recommend — content recommendations related to a specific docs page.
• aws___retrieve_skill — fetch the full content of a domain-specific AWS skill discovered via aws___search_documentation.
• aws___list_regions / aws___get_regional_availability — confirm DSQL or a dependent feature is available in the target region before recommending an architecture.

AWS API tools (require IAM credentials):

• aws___call_aws — execute an authenticated AWS API call. Useful for dsql:CreateCluster, dsql:GetCluster, dsql:ListClusters, etc., when the user wants the assistant to drive cluster lifecycle operations directly. For asynchronous DSQL operations (CreateCluster, DeleteCluster) poll readiness by re-invoking aws___call_aws with dsql:GetCluster — DSQL returns the cluster status directly, not an MCP task ID.
• aws___run_script — sandboxed Python with AWS API access. Useful for multi-step or parallel workflows like “list every cluster in the region, check whose tags include Environment=eval, then describe the matching ones.” May return an MCP task ID for very long scripts.
• aws___get_presigned_url — generate pre-signed Amazon S3 URLs for uploading/downloading files (e.g., DSQL bulk-loading source data).

MCP session tools (no IAM):

• aws___get_tasks — poll MCP-side task IDs returned by aws___call_aws or aws___run_script when the MCP wrapper queues a long-running invocation. NOT for polling AWS-API-side async operations like dsql:CreateCluster — those return their status field directly.

See documentation-tools.md for per-tool detail and example calls.

Setup, auth, and per-assistant invocation differ by client — see the official AWS MCP Server docs (opens in a new tab).

Database Operations

Database operations against a DSQL cluster run through psql by default. The wrapper script scripts/psql-connect.sh handles IAM auth token generation, TLS defaults, application_name tagging, and single-statement guards.

See database-tools.md for the full read / write / schema-discovery patterns.

Detailed References

• input-validation.md — MUST load before building any query. Build SQL with safe_query.build(), which rejects raw strings by construction.
• safe_query.py — the validated-query helper module.
• database-tools.md — psql-based read / write / schema patterns.
• workflow-patterns.md — common multi-step workflow patterns.

Additional Resources

• Aurora DSQL Documentation (opens in a new tab)
• Aurora DSQL Connectivity Tools (opens in a new tab)
• AWS MCP Server (opens in a new tab)