Supports json response format for OpenAI#1367
Merged
Merged
Conversation
Contributor
PR Summary by QodoAdd configurable JSON response format support for OpenAI Description
Diagram
High-Level Assessment
Files changed (5)
|
Contributor
Code Review by Qodo
1. GetChatResponseFormat uses ToLower()
|
![qodo-code-review[bot]](https://avatars.githubusercontent.com/in/484649?s=60&v=4){kind=small}
Comment on lines
+639
to
+644
| return format?.ToLower() switch | ||
| { | ||
| "json" or "json_object" => ChatResponseFormat.CreateJsonObjectFormat(), | ||
| "text" => ChatResponseFormat.CreateTextFormat(), | ||
| _ => null | ||
| }; |
Contributor
There was a problem hiding this comment.
1. getchatresponseformat uses tolower() 📘 Rule violation ≡ Correctness
The new response-format parsing normalizes with format?.ToLower(), which is culture-sensitive and can cause incorrect identifier matching in some locales. This violates the requirement to use ordinal, case-insensitive comparisons for identifier-like strings.
Agent Prompt
## Issue description
The response-format parsing currently normalizes `agent.LlmConfig?.ResponseFormat` using `format?.ToLower()` before switching/matching, which performs a culture-sensitive conversion and can lead to incorrect identifier matching in certain locales. Update this logic to use ordinal, case-insensitive comparisons suitable for identifier-like strings.
## Issue Context
This parsing logic interprets `agent.LlmConfig?.ResponseFormat` values such as `json` and `text`.
## Fix Focus Areas
- src/Plugins/BotSharp.Plugin.OpenAI/Providers/Chat/ChatCompletionProvider.Chat.cs[637-644]
- src/Plugins/BotSharp.Plugin.OpenAI/Providers/Chat/ChatCompletionProvider.Response.cs[593-601]
ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools
Comment on lines
+7
to
+13
| public AgentLlmConfig(AgentTemplateConfig templateConfig) | ||
| { | ||
| Provider = templateLlmConfig.Provider; | ||
| Model = templateLlmConfig.Model; | ||
| MaxOutputTokens = templateLlmConfig.MaxOutputTokens; | ||
| ReasoningEffortLevel = templateLlmConfig.ReasoningEffortLevel; | ||
| Provider = templateConfig.LlmConfig?.Provider; | ||
| Model = templateConfig.LlmConfig?.Model; | ||
| MaxOutputTokens = templateConfig.LlmConfig?.MaxOutputTokens; | ||
| ReasoningEffortLevel = templateConfig.LlmConfig?.ReasoningEffortLevel; | ||
| ResponseFormat = templateConfig.ResponseFormat; |
Contributor
There was a problem hiding this comment.
2. Template drops response_format 🐞 Bug ≡ Correctness
When a template has a valid LlmConfig, InstructService replaces the agent’s LlmConfig with new AgentLlmConfig(template), and the new constructor sets ResponseFormat only from template.ResponseFormat. If the template omits response_format but the agent-level LlmConfig.ResponseFormat is set (e.g., "json"), OpenAI requests won’t enable JSON mode for template-based executions.
Agent Prompt
## Issue description
Template-based LLM overrides currently replace the entire `AgentLlmConfig` and copy `ResponseFormat` only from the template. This causes agent-level `LlmConfig.ResponseFormat` (e.g., `json`) to be lost whenever the template has `LlmConfig.IsValid == true` but does not explicitly set `ResponseFormat`.
## Issue Context
- `InstructService.Execute` and `BuildInnerAgent` create a fresh `AgentLlmConfig` from the template when template `LlmConfig` is valid.
- OpenAI providers read `agent.LlmConfig.ResponseFormat` to set `ChatCompletionOptions.ResponseFormat` / `CreateResponseOptions.TextOptions.TextFormat`.
## Fix Focus Areas
- src/Infrastructure/BotSharp.Abstraction/Agents/Models/AgentLlmConfig.cs[7-14]
- src/Infrastructure/BotSharp.Core/Instructs/Services/InstructService.Execute.cs[246-256]
- src/Infrastructure/BotSharp.Core/Instructs/Services/InstructService.Instruct.cs[67-75]
## Implementation direction
Ensure `ResponseFormat` (and ideally other non-overridden LLM settings) fall back to the agent-level config when the template does not specify them. For example:
- Build `llmConfig` by cloning/starting from `agent.LlmConfig`, then override only the fields present in `template.LlmConfig` and `template.ResponseFormat`.
- Or update the constructor usage so that `ResponseFormat = template.ResponseFormat ?? agent.LlmConfig?.ResponseFormat` (requires passing the base config or doing the merge at the call site).
ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.