creating-oracle-to-postgres-master-migration-plan

creating-oracle-to-postgres-master-migration-plan analyzes a .NET solution and produces a persistent master migration plan for moving from Oracle to PostgreSQL. It discovers every project in the solution, classifies each project for migration eligibility, and writes a structured plan file that downstream agents and skills can parse. It is intended for the start of a multi-project Oracle-to-PostgreSQL migration, for building a migration inventory, or for assessing which .NET projects contain Oracle dependencies.

The workflow runs in four steps. Step 1 discovers projects by finding the solution file (a .sln or .slnx extension, asking the user if multiple exist) and parsing it to extract all .csproj references, noting each project's name, path, and type. Step 2 classifies every non-test project by scanning for Oracle indicators: NuGet references such as Oracle.ManagedDataAccess and Oracle.EntityFrameworkCore (in .csproj and packages.config), Oracle connection strings in appsettings.json, web.config, or app.config, code usage of OracleConnection, OracleCommand, and OracleDataReader, and DDL cross-references under .github/oracle-to-postgres-migration/DDL/Oracle/. Each project receives one of four classifications: MIGRATE, SKIP, ALREADY_MIGRATED, or TEST_PROJECT. Step 3 presents the classified list to the user for adjustments to classifications or migration ordering. Step 4 writes the plan to .github/oracle-to-postgres-migration/Reports/Master Migration Plan.md using an exact template with a solution summary table, a project inventory table, and a migration order section, ordering foundational and shared libraries before their dependents.

The skill targets engineers and teams running .NET-based Oracle-to-PostgreSQL migrations who need a reliable inventory and sequencing plan before conversion work begins. Its output is deliberately structured because downstream consumers depend on the template layout, so following the exact format is a requirement. It is documented in a single SKILL.md and focuses on discovery, classification, and planning rather than performing the actual schema or code conversion.