Update OpenAI library to v6.6.0 and change default model to gpt-5-mini

[Copilot]

Update OpenAI library and change default model to gpt-5-mini

• Update OpenAI library from version 4.80.0 to latest (6.6.0)
• Change DEFAULT_MODEL in src/story.ts from "gpt-4o" to "gpt-5-mini"
• Change default chatModel in src/verify.ts from "gpt-4-turbo" to "gpt-5-mini"
• Change default chatModel in src/chat.ts from "gpt-4-turbo" to "gpt-5-mini"
• Use omni-moderation-latest for automatic updates
• Fix temperature handling for gpt-5-mini (model only supports default temperature)
• Fix TypeScript compatibility issues with OpenAI v6 API changes
• Add comprehensive JSDoc documentation to all exported methods
• Update README with detailed API reference and usage examples
• Run lint and build to verify changes
• Run security scan (no vulnerabilities found)

Summary

All requested changes have been successfully implemented with fixes for API compatibility and comprehensive documentation:

1. OpenAI Library Update: Updated from v4.80.0 to v6.6.0 (latest)
2. Model Changes: Updated all default models to gpt-5-mini
3. Moderation Model Update: Using omni-moderation-latest to automatically use the latest version without requiring updates
4. Temperature Fix: gpt-5-mini only supports default temperature (1), so temperature parameter is now conditionally omitted for this model
5. API Compatibility: Fixed TypeScript type issues introduced by OpenAI v6 API changes
6. Documentation: Added comprehensive JSDoc comments to all exported classes and methods (Story, ChatAssistant, ImageGenerator, verifyPrompt), and updated README with detailed API reference including parameters, return types, and usage examples

All builds and linters pass successfully.

Fixes #14

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Update to latest</issue_title>
<issue_description>Update the OpenAI library to latest and ensure that the model used by default is gpt-5-mini.

You will need to update the openai library.

This needs to be changed in:

• https://github.com/CodingBull-dev/story-gpt/blob/main/src/story.ts#L8
• https://github.com/CodingBull-dev/story-gpt/blob/main/src/verify.ts#L17
• https://github.com/CodingBull-dev/story-gpt/blob/main/src/chat.ts#L14

Also update the omni-moderation-latest in src/verify.ts to omni-moderation.</issue_description>

Comments on the Issue (you are @copilot in this section)

Fixes #14

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

[coderabbitai]

Important

Review skipped

Bot user detected.

To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Note

Other AI code review bot(s) detected

CodeRabbit has detected other AI code review bot(s) in this pull request and will avoid duplicating their findings in the review comments. This may lead to a less comprehensive review.

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[Bullrich]

@copilot please fix the test. You can try to run them first to see if they even get to work (even if it fails because you are missing an api key).

This are the errors:

Run node e2e-test.mjs
🚀 Starting E2E tests for story-gpt
📝 Test 1: Testing createStory()
Generating story for prompt A short tale about a brave mouse
❌ createStory failed: 400 Unsupported value: 'temperature' does not support 0.44 with this model. Only the default (1) value is supported.
📝 Test 2: Testing Story.generateStory()
Generating story for prompt A tiny story about a cat
❌ Story.generateStory failed: 400 Unsupported value: 'temperature' does not support 0.4 with this model. Only the default (1) value is supported.
📝 Test 3: Testing verifyPrompt() with valid prompt
❌ verifyPrompt (valid) failed: 400 Invalid value for 'model' = omni-moderation. Please check the OpenAI documentation and try again.
📝 Test 4: Testing verifyPrompt() with invalid prompt
❌ verifyPrompt (invalid) failed: 400 Invalid value for 'model' = omni-moderation. Please check the OpenAI documentation and try again.
📝 Test 5: Testing ImageGenerator.generateImage()

[Copilot]

@copilot please fix the test. You can try to run them first to see if they even get to work (even if it fails because you are missing an api key).

This are the errors:
...

Fixed in commit 0423b1e.

The tests were failing due to two issues:

1. Temperature: gpt-5-mini only supports the default temperature value (1). Updated ChatAssistant to conditionally omit the temperature parameter when using this model.
2. Moderation model: omni-moderation is not a valid model name in the OpenAI API. Changed to omni-moderation-2024-09-26 which is the current stable version.

The build now passes successfully.

[Bullrich]

@copilot please address my comments and also let me know, have you updated the docs for the values that now are missing if there are any docs? If not, please document the methods that we expose

[Bullrich]

can we please use latest or whatever is available that means something similar? I don't want to use dated versions that end up expiring.

[Copilot]

Changed to omni-moderation-latest in commit f4ed18e. This will automatically use the latest moderation model without requiring updates when new versions are released.

[Copilot]

@copilot please address my comments and also let me know, have you updated the docs for the values that now are missing if there are any docs? If not, please document the methods that we expose

Addressed in commit f4ed18e:

1. Changed moderation model to omni-moderation-latest
2. Added comprehensive JSDoc documentation to all exported methods in:
   • Story class (constructor, generateStory, generateTitle, generateImage)
   • ChatAssistant class (constructor, chat method)
   • ImageGenerator class (constructor, generateImage, generateImages)
   • verifyPrompt function
3. Updated README with complete API reference including:
   • All exported functions and classes
   • Parameter descriptions and types
   • Return types
   • Usage examples
   • Important notes about temperature with gpt-5-mini

All JSDoc comments now correctly reference gpt-5-mini as the default model and note that it only supports temperature=1.