Issues

• [Bug]: a failing non-FunctionTool tool runs twice through call_tool's kwargs fallback

  run-llama/llama_index#23461 · Filter by author chrikrah opened on Oct 11, 2026
  1

• [Bug]: Anthropic stream_chat: server-tool input_json_delta after a completed client tool_use block overwrites the finished client tool call's arguments

  run-llama/llama_index#23459 · Filter by author BlueX888 opened on Oct 11, 2026
  1

• [Bug]: Refine with streaming=True and output_cls returns "Empty Response"

  run-llama/llama_index#23446 · Filter by author digit50 opened on Oct 10, 2026

• [Bug]: LLM.predict_and_call raises ValidationError without react_chat_formatter

  run-llama/llama_index#23445 · Filter by author digit50 opened on Oct 10, 2026

• [Bug]: HierarchicalNodeParser async methods return the documents unsplit

  run-llama/llama_index#23444 · Filter by author digit50 opened on Oct 10, 2026

• [Bug]: QdrantVectorStore numeric EQ/NE filters depend on int vs float and miss float payloads

  bugtriage
  run-llama/llama_index#23440 · Filter by author skolchin opened on Oct 9, 2026

• [Bug]: CodeActAgent drops assistant tool-call metadata during handoff, leaving an orphan tool response

  run-llama/llama_index#23438 · Filter by author dakjdakd opened on Oct 9, 2026

• [Bug]: Property graph relation keys collide when entity IDs or labels contain underscores

  run-llama/llama_index#23436 · Filter by author dakjdakd opened on Oct 9, 2026

• [Bug]: Pinecone document deletion fallback can delete other documents with overlapping IDs

  run-llama/llama_index#23434 · Filter by author dakjdakd opened on Oct 9, 2026

• [Bug]: RecursiveRetriever incorrectly converts zero scores to 1.0

  bugtriage
  run-llama/llama_index#23429 · Filter by author Mohit-25-tech opened on Oct 9, 2026
  1

• [Bug]: MarkdownNodeParser misses headings indented by 1–3 spaces and loses their hierarchy

  run-llama/llama_index#23411 · Filter by author Ra1nhy opened on Oct 7, 2026
  1

• [Bug]: Support for GPT-6.1 Sol in llama-index-llms-openai

  enhancementtriage
  run-llama/llama_index#23404 · Filter by author laslil7437 opened on Oct 6, 2026

• [Bug]: vLLM notebook fails with unsupported swap_space and best_of arguments

  bugtriage
  run-llama/llama_index#23399 · Filter by author SjeYinTeoIntel opened on Oct 6, 2026

• [Bug]: parse_json_markdown loses literal JSON fence markers inside valid payloads

  run-llama/llama_index#23391 · Filter by author hulkbig opened on Oct 5, 2026

• [Bug]: md_to_df adds a column or returns None with CRLF/trailing row whitespace

  run-llama/llama_index#23376 · Filter by author hulkbig opened on Oct 5, 2026

• [Documentation]: Clarify discovery and contribution route for external tools

  run-llama/llama_index#23372 · Filter by author zealous1 opened on Oct 5, 2026
  2

• [Bug]: BedrockConverse can't use OpenAI GPT-6.1 Sol or Claude Fable 5.1 (model IDs rejected; Fable 5.1 also rejects forced toolChoice)

  run-llama/llama_index#23366 · Filter by author kimnamu opened on Oct 5, 2026

• [Bug]: KeywordTable.node_ids raises on an empty keyword index

  run-llama/llama_index#23359 · Filter by author imharjot opened on Oct 4, 2026
  1

• [Bug]: reciprocal_rerank and simple fusion still mutate retriever-owned score wrappers

  run-llama/llama_index#23351 · Filter by author eddyflores100-lang opened on Oct 3, 2026
  2

• [Bug]: RankLLMRerank cannot be imported without vllm installed, regardless of reranking backend

  buggood first issuetriage
  run-llama/llama_index#23339 · Filter by author kostasrigatos opened on Oct 2, 2026
  2

• [Bug]: SimpleFileNodeParser ignores custom id_func for supported file types

  run-llama/llama_index#23334 · Filter by author takagibit18 opened on Oct 2, 2026

• [Bug]: Relative score fusion corrupts rankings when cached results share score wrappers

  run-llama/llama_index#23332 · Filter by author michaelswissa opened on Oct 2, 2026

• Integration example proposal: typed bounded retrieval over large MCP tool sets

  run-llama/llama_index#23331 · Filter by author JDeun opened on Oct 2, 2026

• [Bug]: RetrieverEvaluator, MultiModalRetrieverEvaluator and ObjectRetriever.aretrieve still call sync postprocess_nodes() in async paths

  run-llama/llama_index#23326 · Filter by author Zhuoxi2000 opened on Oct 1, 2026
  1

• [Bug]: MarkdownNodeParser treats more than six opening hashes as a heading

  bugtriage
  run-llama/llama_index#23322 · Filter by author Marcuswang0824 opened on Oct 1, 2026
  1