Issues

• AG-UI MessagesSnapshot: str() coercion corrupts message content (None -> 'None', multimodal -> pydantic repr)

  vibrantlabsai/ragas#3060 · Filter by author contextablemark opened on Oct 11, 2026
  1

• FewShotPydanticPrompt never forwards threshold_for_examples to ExampleStore.get_examples — DemonstrationConfig.threshold silently ignored (store default 0.7 always used)

  vibrantlabsai/ragas#3056 · Filter by author BlueX888 opened on Oct 11, 2026
  1

• This project is not maintained anymore

  question
  vibrantlabsai/ragas#3053 · Filter by author kubilaycaglayan opened on Oct 9, 2026

• Integration proposal: docs page for using Ragas with hieevas (agent process & safety metrics)

  enhancement
  vibrantlabsai/ragas#3051 · Filter by author sampath-6 opened on Oct 5, 2026

• New metric: structural filter-correctness for metadata-filtered retrieval

  enhancement
  vibrantlabsai/ragas#3045 · Filter by author ghubuser-dnt opened on Sep 30, 2026
  5

• answer_relevancy blocks the entire event loop via synchronous embed_query/embed_documents, stalling unrelated concurrent tasks under evaluate()

  bug
  vibrantlabsai/ragas#3040 · Filter by author ecalogero opened on Sep 27, 2026

• griptape integration: ImportError raised when griptape is missing tells the user to install opik

  vibrantlabsai/ragas#3032 · Filter by author BlueX888 opened on Sep 25, 2026
  1

• llm_factory: OpenAI GPT-6 Sol/Luna/Astra on Amazon Bedrock fail with 400 because Bedrock model IDs (openai.gpt-6-sol) skip the GPT-5+ parameter mapping

  vibrantlabsai/ragas#3031 · Filter by author kimnamu opened on Sep 24, 2026
  1

• [Feature] First-class decision model abstraction for bounded metric judgments

  enhancement
  vibrantlabsai/ragas#3030 · Filter by author sanatb187 opened on Sep 23, 2026
  5

• A row that errors leaves the denominator, so the reported score rises as the system degrades

  vibrantlabsai/ragas#3028 · Filter by author shaurya416 opened on Sep 23, 2026
  3

• llm_factory: mode= is silently discarded for provider='anthropic' and provider='google'/'gemini' (new google-genai)

  vibrantlabsai/ragas#3021 · Filter by author BlueX888 opened on Sep 19, 2026
  1

• AgentGoalAccuracyWithoutReference never sets output_type, so metric.train() instruction optimization aborts with ValueError

  vibrantlabsai/ragas#3014 · Filter by author BlueX888 opened on Sep 16, 2026
  1

• Inconsistency between claims and statements in context_recall

  question
  vibrantlabsai/ragas#3013 · Filter by author dosnshcu opened on Sep 16, 2026
  2

• AG-UI MessagesSnapshot: convert_messages_snapshot drops tool call name and arguments (ToolCall(name='unknown_tool', args={}))

  vibrantlabsai/ragas#3010 · Filter by author BlueX888 opened on Sep 15, 2026

• # Question: Recommended integration approach for Node.js / TypeScript full-stack applications

  question
  vibrantlabsai/ragas#3009 · Filter by author Altria1979 opened on Sep 14, 2026

• Is this project actively maintained at this point?

  question
  vibrantlabsai/ragas#3004 · Filter by author arpieb opened on Sep 10, 2026
  3

• New metric: PII Leakage detection for RAG/agentic responses

  enhancement
  vibrantlabsai/ragas#2998 · Filter by author rajeshl8 opened on Sep 8, 2026
  1

• ModuleNotFoundError: langchain_community.chat_models.vertexai (breaks import ragas)

  vibrantlabsai/ragas#2995 · Filter by author Sahilll15 opened on Sep 4, 2026
  1

• Metric integrity: faithfulness and context-recall divide by emitted verdicts not submitted inputs; errors become NaN and exit headline; context embedded unsanitized into judge prompts; extract_json takes first structure

  vibrantlabsai/ragas#2992 · Filter by author AUTHENSOR opened on Sep 3, 2026
  2

• CI: check-docs fails on every fork PR — pull_request_target checkout blocked by actions/checkout fork guard

  vibrantlabsai/ragas#2989 · Filter by author linhongyu510 opened on Sep 2, 2026

• 使用 context manager 修复文件资源泄漏

  bug
  vibrantlabsai/ragas#2986 · Filter by author YutShang opened on Sep 2, 2026
  1

• LLM cache keys can return cross-model responses

  vibrantlabsai/ragas#2983 · Filter by author YusefSyed opened on Sep 2, 2026
  2

• [Bug]: Faithfulness metric (and PydanticPrompt) fails to parse JSON when LangChain LLM response contains markdown code fences

  bug
  vibrantlabsai/ragas#2981 · Filter by author hyolee1999 opened on Sep 1, 2026

• Should the reported metric mean disclose how many samples it was computed over?

  vibrantlabsai/ragas#2980 · Filter by author roy-tong opened on Sep 1, 2026
  1

• OpenAI client base_url /v1 plus Instructor adapter for a Chat Completions host

  vibrantlabsai/ragas#2978 · Filter by author cursor opened on Aug 28, 2026