Dup Classes
Testing & Qualitydup-classesDup Classes is an Agent Skill for testing & quality published in the Azure/azure-sdk-for-java repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.
Verify whether generated Java classes duplicate openai-java models by comparing fields/types (names may differ). Use when checking for duplicate model coverage.
How to use this skill
Bring this guide into your coding agent with a prompt tailored to the tool you use.
- Open your project in Codex.
- Copy and paste the prompt below.
- Review files and risks before approving any action.
Inspect the Agent Skill at https://github.com/Azure/azure-sdk-for-java/blob/HEAD/sdk/ai/.workflow_docs/dup-classes/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/dup-classes, but wait for my approval before writing files or running scripts.
Copying this prompt does not install or run the skill. Review third-party files before use.
Duplicate Class Verification (Generated vs openai-java)
Use this skill to compare generated Java models against the openai-java dependency. The goal is field-by-field comparison of model shapes, even when class or field names differ.
Inputs to confirm
- Generated source root (e.g.,
src/main/java/...) - The relevant
pom.xml(module) to resolve theopenai-javadependency version - Optional: package or class name hints to narrow the search
Steps
-
Locate the pom.xml in the current directory tree (
find . -name pom.xml). If multiple, ask which module to use. -
Resolve openai-java:
- Search the chosen pom for
openai-java(or an explicit group/artifact provided by the user). - Resolve the version (including properties) and locate the JAR in
~/.m2/repository. - The model classes live in
openai-java-core, not the top-levelopenai-javaartifact.
- Search the chosen pom for
-
List candidate classes:
- Generated classes: scan the source root for
classandrecorddeclarations. - openai-java classes:
jar tf <jar> | grep '\.class$'(filter by package hints if provided).
- Generated classes: scan the source root for
-
Extract field signatures (names may differ; compare shape):
- Generated source:
- For
record, use the component list in therecorddeclaration. - For
class, extract non-static field declarations (type + count) and note any@JsonPropertynames. - Check
toJson/fromJsonmethods for the actual JSON keys used in serialization.
- For
- openai-java JAR:
- Use
javap -classpath <jar> -p <FQCN>to list fields (ignorestatic,validated,hashCode$delegate,additionalProperties). - Extract
@JsonPropertykeys from the sources JAR (*-sources.jar) for JSON key comparison.
- Use
- Generated source:
-
Compare shapes:
- Compare field count and field types (order-independent).
- Compare JSON keys from
@JsonProperty(openai-java) vstoJson/fromJsonstring literals (generated). - Compare enum/union values when field types are enums or string unions.
- Follow type hierarchy:
BinaryData↔JsonValue(both represent untyped JSON),Map<String,BinaryData>↔Map<String,JsonValue>, Java enum ↔ Kotlin string enum.
-
Categorize results (do NOT treat all matches the same):
Actionable duplicates — standalone models not in any type hierarchy. These can potentially be suppressed and replaced with the openai-java equivalent. Examples:
ComparisonFilter,Reasoning.Structural equivalents — classes that produce identical JSON but participate in a discriminator hierarchy (e.g.,
extends Tool,extends TextResponseFormatConfiguration). The SDK's polymorphic serialization (Tool.fromJson()dispatches toFunctionTool.fromJson(), etc.) requires these to exist. They are NOT actionable duplicates. Examples:FunctionTool,FileSearchTool,ComputerUsePreviewTool.Partial matches — classes with most fields matching but extra Azure-specific fields. Note the extra fields. Examples:
CodeInterpreterTool(extracontainer),WebSearchTool(extracustom_search_configuration). -
Report:
- Provide a table with: generated class → openai-java class, field count, matching fields, category.
- Clearly separate actionable duplicates from structural equivalents.
- For actionable duplicates, note whether
@@alternateTypeor@@access(internal)would be the right suppression mechanism (seededup-openaiskill).
Useful commands
List generated class names
rg -n "^(public\s+)?(final\s+)?(class|record)\s+" <generated_root>
Extract field lines from source (classes)
grep -E '^\s+private\s+' <file> | grep -v 'static\s'
Extract JSON keys from generated toJson/fromJson
grep -E 'jsonWriter\.write|"[a-z_]+"' <file> | grep -v '//'
Inspect fields in a JAR class
javap -classpath <jar> -p <fully.qualified.ClassName>
Extract @JsonProperty from openai-java sources JAR
jar xf <sources-jar> main/com/openai/models/<Class>.kt
grep '@JsonProperty' main/com/openai/models/<Class>.kt
Check if a class participates in a hierarchy
grep 'extends\s' <file> # If it extends Tool, TextResponseFormatConfiguration, etc. → structural
Notes
- Use
search-m2if you need help locating the dependency version or JAR path. - If the user provides only a vague class hint, narrow candidates by package or field count first.
- The openai-java classes are Kotlin and use Jackson; generated classes use azure-json (
JsonSerializable). Compare at the JSON wire level, not at the Java API level.