[bot] RubyLLM moderation API (`RubyLLM.moderate`) not instrumented
还没有人认领这个 Issue。
评估
- 难度
- 3/5
- 预计耗时
- 1-2 天
- 新手友好度
- 72/100
- Issue 类型
- 功能
- 描述清晰度
- 描述清楚
- 活跃度
- 冷清
- 技术栈
- ruby
调研方向
从 lib/braintrust/contrib/ruby_llm/patcher.rb 和 integration.rb 开始，然后对比 lib/braintrust/contrib/openai/instrumentation/moderations.rb 以及 openai/patcher.rb 第 131–170 行中的 moderation 模式。完成标准是：RubyLLM.moderate 及其 alias 已使用指定的输入、元数据、moderation 输出和延迟进行 instrument，integration 已注册，并且加载检测已更新。
由索引模型根据 Issue 内容生成。
描述
Summary
RubyLLM.moderate is a model execution call that invokes OpenAI's moderation API to classify text for harmful content. It is not instrumented by the Braintrust Ruby SDK. The existing RubyLLM integration only patches RubyLLM::Chat.
This is distinct from issue #136 (RubyLLM embed, paint, and transcribe APIs not instrumented), which covers three other non-chat RubyLLM execution surfaces but explicitly omits moderation.
What is missing
RubyLLM.moderate("text") (aliased as RubyLLM::Moderation.moderate) makes a live API request to OpenAI's moderation endpoint using models such as omni-moderation-latest or text-moderation-007. It returns a RubyLLM::Moderation object with:
flagged— overall boolean- Per-category flags and confidence scores
idandmodelfrom the API response
This is the same class of execution call as the moderations already instrumented for the openai and ruby-openai integrations (ModerationsPatcher in both). A parallel RubyLLM::ModerationPatcher would follow the same pattern.
What a span should capture
- Input: the text string being moderated
- Metadata: model (e.g.,
omni-moderation-latest), provider (ruby_llm), endpoint - Output: flagged categories and scores (or summary)
- Metrics: no token usage is returned by the moderation API, but latency should be captured
Braintrust docs status
not_found — The Braintrust RubyLLM integration docs at https://www.braintrust.dev/docs/integrations/sdk-integrations/ruby-llm cover chat completions and tool calls but do not mention moderation. The OpenAI provider docs at https://www.braintrust.dev/docs/providers/openai document moderation instrumentation via the openai gem but not via RubyLLM.
Upstream sources
- RubyLLM moderation docs: https://rubyllm.com/moderation/ — documents
RubyLLM.moderate("text")making an API call to OpenAI'somni-moderation-latestmodel - RubyLLM README: https://github.com/crmne/ruby_llm — lists
RubyLLM.moderateunder core capabilities
Local files inspected
lib/braintrust/contrib/ruby_llm/patcher.rb— defines onlyChatPatcher; noModerationPatcherlib/braintrust/contrib/ruby_llm/integration.rb— registers only[ChatPatcher];loaded?checksdefined?(::RubyLLM::Chat)onlylib/braintrust/contrib/openai/instrumentation/moderations.rb— demonstrates the existing moderation instrumentation pattern in this repolib/braintrust/contrib/openai/patcher.rb(lines 131–170) —ModerationsPatcherfor reference
- 主要语言
- Ruby
- 星标
- 9
- 派生
- 11
- 平均合并
- 2 天 3 小时
- 30 天内合并 PR
- 5
环境准备
- 提供 Dockerfile 或 Docker Compose 文件
- 没有 Pull Request 模板
- 阅读贡献指南
从这里开始
- 先读完整个 Issue，再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库，在一个分支上完成修改。
- 提交 Pull Request，并在描述里引用这个 Issue 编号。
braintrustdata/braintrust-sdk-ruby 的其他 Issue
-
难度 2/5 1-3 小时 新手友好度 78/100
-
难度 2/5 1-3 小时 新手友好度 76/100
-
难度 2/5 1-3 小时 新手友好度 72/100
-
ruby
难度 3/5 半天 新手友好度 62/100
-
ruby
难度 3/5 1-2 天 新手友好度 45/100
查看 braintrustdata/braintrust-sdk-ruby 的全部 Issue
相似的 Issue
-
难度 1/5 1 小时以内 新手友好度 84/100
-
难度 2/5 1-3 小时 新手友好度 65/100
-
难度 2/5 1-3 小时 新手友好度 67/100
-
难度 1/5 1 小时以内 新手友好度 88/100
TheOdinProject/curriculum#31487 · 2 条评论 ·
维护者通常 1 天内回复
-
Bug: Assistant replies lose the space before numbers ("Oct7", "as of10-22-2026") with OpenAI models未关闭
难度 2/5 1-3 小时 新手友好度 84/100
we-promise/sure#4441 ·
维护者通常 1 天内回复