Database Architect — DB Design Full Pipeline

Workflow

Phase 1: Preparation (Performed directly by the orchestrator)

1. Extract from user input:
   • Domain: What service is the DB for
   • DBMS: PostgreSQL / MySQL / MongoDB / DynamoDB
   • Core Entities: Primary data subjects
   • Expected Scale (optional): Row counts, TPS
   • Existing Files (optional): Existing schemas, ERDs, SQL, etc.
2. Create the _workspace/ directory at the project root
3. Organize the input and save to _workspace/00_input.md
4. If existing files are provided, copy them to _workspace/ and skip the corresponding phase
5. Determine the execution mode based on the scope of the request

Phase 2: Team Assembly and Execution

Tasks 3a (performance) and 3b (security) are executed in parallel.

Inter-team communication flow:

• data-modeler completes -> Delivers DDL basis to migration-manager, access patterns to performance-analyst, sensitive data to security-auditor
• migration-manager completes -> Delivers index DDL to performance-analyst, permission DDL to security-auditor
• performance-analyst <-> security-auditor: Mutually verify that performance optimizations do not compromise security
• integration-reviewer cross-validates all artifacts. When 🔴 must-fix issues are found, requests revisions from the relevant agent -> rework -> re-verify (up to 2 rounds)

Phase 3: Integration and Final Artifacts

Organize the final artifacts based on the review report:

1. Verify all files in _workspace/
2. Confirm that all 🔴 must-fix items from the review report have been addressed
3. Report the final summary to the user

Mode by Task Scale

Leveraging existing files: If the user provides schemas, ERDs, or other existing files, skip the corresponding steps.

Data Transfer Protocol

File naming convention: {order}_{agent}_{artifact}.{extension}

Error Handling

Test Scenarios

Normal Flow

Prompt: "Design a PostgreSQL database for an e-commerce platform. I need user, product, order, payment, and review tables. Expecting 100K orders per day" Expected Result:

• Model: 5 core tables + junction tables, 3NF normalization, ERD
• Migration: Sequential DDL + rollback scripts + seed data
• Performance: Index strategy, key query optimization, partitioning design
• Security: RBAC, PII encryption, audit logging, backup strategy
• Review: All items in the alignment matrix verified

Existing File Flow

Prompt: "Optimize the performance of this SQL schema" + SQL file Expected Result:

• Copy existing schema to _workspace/02_migration.sql
• Optimization mode: deploy performance-analyst + integration-reviewer
• Skip data-modeler, migration-manager, security-auditor

Error Flow

Prompt: "Design a database, blog platform" Expected Result:

• Scale/DBMS unknown -> data-modeler infers PostgreSQL + standard blog entities (Post, User, Comment, Tag)
• Execute in full pipeline mode
• Review report notes "design based on inferred requirements"

Agent Extension Skills

Extension skills that enhance each agent's domain expertise: