Files
awesome-copilot/skills/migrating-oracle-to-postgres-data-access-code/SKILL.md
T
Paul Delannoy f7e8aaa2d8 'Oracle-to-PostgreSQL Migration Expert' Custom Agent: Update Custom Agent & Plugin & Skills (#2566)
* Enhance Oracle-to-PostgreSQL migration skills and documentation

- Update migration agent guidelines to prioritize extension tool usage for code migration.
- Refine migration phases with detailed steps for pre-migration review and schema migration.
- Add new reviewing skill references for PostgreSQL materialized view refresh and UNION ALL planner risks.
- Ensure consistency in collation handling and testing strategies across skills.

* - Review migration phases to ensure correct order of execution
- Add exit criteria to each migration phase
- Remove invocation of `ms-ossdata.vscode-pgsql` extension due to dependency on VSCode
- Update README.md to reflect changes in migration phases and exit criteria
- Update broken reference to custom agent in plugin.json

* Enhance downstream migration skills and following custom agent improvements

- Added new skill for migrating .NET data access code from Oracle to PostgreSQL.
- Updated README to include new migration skill.
- Improved existing skills with clearer file naming conventions and migration actions.
- Added reference documents for handling Oracle-specific functions and pagination.
- Created detailed guides for NVL, DECODE, ROWNUM, SYSDATE, and DUAL replacements.

* Clarify PostgreSQL schema immutability and stored procedure migration risks

* Update target schema path in migration documentation for clarity

* fix(skills): clarify Phase 3-only scope for Oracle test skills

Both Oracle test skills were ambiguously worded in ways that could
cause a model to invoke them during Phase 6 (PostgreSQL test migration)
instead of using them exclusively in Phase 3.

Key changes:
- Rewrite descriptions to explicitly state Phase 3-only usage and
  warn against invoking during Phase 6
- Replace "scaffold for Oracle first" with "Oracle only" to remove
  the false implication of a second PostgreSQL scaffolding step
- Replace "Tests validate behavior consistency when running against
  Oracle or PostgreSQL" with clear Oracle-only framing
- Rename "DB-agnostic assertions" → "Assertion portability" and
  explain the why (survive Phase 6 migration without rewrites)
- Fix datetime bullet in integration tests skill to use generic
  Oracle column precision language instead of PostgreSQL type syntax
- Name Oracle NuGet package explicitly (Oracle.ManagedDataAccess.Core)

* feat(oracle-to-postgres): gate Phase 1 on DDL presence; add DDL scan to Phase 2 risk analysis

- Phase 1 success criteria now requires Oracle DDL artifacts to be
  confirmed present at the recorded location before proceeding.
  If missing, the agent stops and prompts the user to provide them.

- Phase 2 risk analysis now explicitly scans DDL/Oracle/{ProjectName}/
  as supplemental context, summarising procedure complexity indicators
  (dynamic SQL, DBMS_* / UTL_* references, autonomous transactions,
  pipelined functions, BULK COLLECT/FORALL, REF CURSOR, TYPE bodies)
  rather than ingesting DDL files wholesale. This ensures schema-level
  migration risk is captured even when it isn't visible in application
  code alone.

* - Merge in latest 'main' changes
- Update custom agent plugin (eg resolve conflict and add new skill)
- Validate skills
- Run build

* chore(plugin.json): update version to 1.1.0

* feat(oracle-to-postgres): update version to 1.1.0 for migration expert plugin

* fix: add INOUT to ignore-words-list for PostgreSQL migration

---------

Co-authored-by: TCPrimedPaul <paul.delannoy@tc.gc.ca>
2026-08-11 12:47:51 +10:00

7.9 KiB

name, description
name description
migrating-oracle-to-postgres-data-access-code Migrates .NET/C# data access code from Oracle to PostgreSQL (Npgsql). Replaces Oracle NuGet packages, rewrites OracleConnection/OracleCommand/OracleDataReader usage, fixes DbType mappings, updates stored procedure invocation patterns, and adapts connection string configuration. Use when migrating the application code layer of a .NET project during an Oracle-to-PostgreSQL database migration.

Migrating .NET Data Access Code from Oracle to PostgreSQL

Migrate the C# data access layer of a single .Postgres-copy project from Oracle (Oracle.ManagedDataAccess) to PostgreSQL (Npgsql). Work item by item through Reports/{ProjectName}/MigrationChecklist.md.

Prerequisites

  • The .Postgres project copy exists (created in Phase 5 setup).
  • Reports/{ProjectName}/MigrationChecklist.md exists and is the source of truth for what to change.
  • Reports/{ProjectName}/OracleRiskAnalysis.md exists for cross-referencing behavioral differences.

Workflow

Progress:
- [ ] Step 1: Replace NuGet packages
- [ ] Step 2: Update connection string configuration
- [ ] Step 3: Rewrite ADO.NET type references
- [ ] Step 4: Fix DbType mappings
- [ ] Step 5: Migrate stored procedure invocation
- [ ] Step 6: Address Oracle-specific SQL and syntax
- [ ] Step 7: Build and verify

Step 1: Replace NuGet packages

In the .csproj of the .Postgres project:

  • Remove: Oracle.ManagedDataAccess.Core, Oracle.EntityFrameworkCore (and any other Oracle.* packages)
  • Add: Npgsql (for ADO.NET) and/or Npgsql.EntityFrameworkCore.PostgreSQL (for EF Core)
  • Keep version pinning consistent with the target .NET version; do not introduce newer package versions than what the solution already uses for similar packages.
  • If System.Data abstractions (IDbConnection, IDbCommand) are used project-wide, the surface-level code may need fewer changes — identify them first.

Step 2: Update connection string configuration

  • Locate the Oracle connection string in appsettings.json, appsettings.{env}.json, web.config, app.config, or environment variable configuration.
  • Replace with a Npgsql-compatible connection string: Host=localhost;Port=5432;Database=mydb;Username=myuser;Password=mypassword
  • Do not hardcode credentials — use the same configuration mechanism already in use (e.g., environment variables, secrets manager, IConfiguration).
  • Update any named connection string keys only if they were Oracle-specific (e.g., OracleConnection). Prefer keeping the same key name to minimize application config changes.

Step 3: Rewrite ADO.NET type references

Replace Oracle-specific ADO.NET types with Npgsql equivalents:

Oracle type Npgsql replacement
OracleConnection NpgsqlConnection
OracleCommand NpgsqlCommand
OracleDataReader NpgsqlDataReader
OracleDataAdapter NpgsqlDataAdapter
OracleParameter NpgsqlParameter
OracleTransaction NpgsqlTransaction
OracleException NpgsqlException
OracleDbType NpgsqlDbType (from NpgsqlTypes namespace)

Update using directives accordingly (using Oracle.ManagedDataAccess.Clientusing Npgsql).

If the codebase uses IDbConnection/IDbCommand abstractions registered via DI, update only the DI registration and connection string — the consuming code may not need changes.

Step 4: Fix DbType and NpgsqlDbType mappings

Oracle parameter types do not map 1:1 to Npgsql. Review every OracleParameter (now NpgsqlParameter) that sets an explicit type:

Oracle type Notes
OracleDbType.Varchar2 Use NpgsqlDbType.Varchar or omit (Npgsql infers from value)
OracleDbType.Clob Use NpgsqlDbType.Text
OracleDbType.Number Use NpgsqlDbType.Numeric or NpgsqlDbType.Integer depending on precision
OracleDbType.Date Use NpgsqlDbType.Date (date only) or NpgsqlDbType.Timestamp (if time component used)
OracleDbType.TimeStamp Use NpgsqlDbType.Timestamp
OracleDbType.RefCursor Use NpgsqlDbType.Refcursor — see Step 5
OracleDbType.Char Use NpgsqlDbType.Char

For parameters where Oracle inferred the type from the value, Npgsql also infers — explicit type setting is often unnecessary and can be removed.

Step 5: Migrate stored procedure invocation

Oracle and PostgreSQL stored procedure invocation differ significantly:

  • Command type: Retain CommandType.StoredProcedure for function calls. For procedures that use OUT parameters, PostgreSQL requires CommandType.Text with CALL proc_name(...) syntax in some versions of Npgsql — verify against the target Npgsql version.
  • RefCursor handling: Oracle returns ref cursors as output parameters; PostgreSQL returns them differently:
    • For RETURNS TABLE / RETURNS SETOF, use ExecuteReader() directly — no cursor parameter needed.
    • For RETURNS refcursor, call within a transaction, read the cursor name from the output parameter, then issue FETCH ALL IN "<cursor_name>".
    • Remove any Oracle-specific cursor-wrapping code (e.g., OracleRefCursor).
  • OUT parameters: PostgreSQL stored procedures use INOUT or function return values. Verify parameter direction matches the migrated procedure signature.
  • Sequence NEXTVAL: Replace SELECT {SEQUENCE}.NEXTVAL FROM DUAL with SELECT nextval('{sequence_name}').
  • Named parameters: Npgsql uses @param_name; Oracle used :param_name. Update all parameter name prefixes.

Step 6: Address Oracle-specific SQL and C# patterns

Review inline SQL strings and query builders for Oracle-specific constructs and replace:

Oracle construct PostgreSQL replacement
ROWNUM <= n LIMIT n
ROWNUM = 1 LIMIT 1
NVL(x, y) COALESCE(x, y)
DECODE(expr, v1, r1, ...) CASE WHEN expr = v1 THEN r1 ... END
SYSDATE / SYSTIMESTAMP NOW() or CURRENT_TIMESTAMP
TO_CHAR(date, fmt) TO_CHAR(date, fmt) (mostly compatible; verify format strings)
TO_DATE(str, fmt) TO_DATE(str, fmt) (verify format strings)
TO_NUMBER(str) CAST(str AS NUMERIC) or str::NUMERIC
`
DUAL table Remove FROM DUAL; PostgreSQL evaluates SELECT expr without a table
CONNECT BY hierarchy Rewrite using recursive CTEs (WITH RECURSIVE)
MERGE INTO Rewrite as INSERT ... ON CONFLICT DO UPDATE
Empty string '' as NULL Oracle treats '' as NULL; PostgreSQL does not — check comparisons and IS NULL guards
VARCHAR2 VARCHAR or TEXT

Step 7: Build and verify

After addressing all checklist items:

  1. Run dotnet build on the .Postgres project. Fix any remaining compilation errors.
  2. Verify no Oracle-specific namespaces remain: search for Oracle.ManagedDataAccess, OracleConnection, OracleCommand, :param patterns.
  3. Mark completed items in Reports/{ProjectName}/MigrationChecklist.md.

EF Core projects

If the project uses Oracle.EntityFrameworkCore:

  • Replace the provider registration in DbContext configuration: .UseOracle(...).UseNpgsql(...)
  • Replace OracleDbContextOptionsBuilder references.
  • Review OnModelCreating for Oracle-specific configurations (e.g., HasColumnType("NUMBER")HasColumnType("numeric")).
  • Sequence configuration: modelBuilder.HasSequence<int>("seq_name").StartsAt(1).IncrementsBy(1) syntax is compatible; verify column defaults referencing sequences.
  • Do not run EF Core migrations — schema is managed externally via DDL scripts (Phase 4).

Key Constraints

  • Work only within the .Postgres copy — never modify the original Oracle-targeting project.
  • Keep to existing .NET and C# versions; do not introduce newer language or runtime features.
  • Preserve comments and application logic; change only what is necessary for PostgreSQL compatibility.
  • Oracle is the source of truth — behavioral differences must be documented as bug reports, not silently altered.