All issues

Search results

[Bug]: llama-index-llms-openai-like blocks llama-index-llms-openai 0.8.x

bugSomething isn't workingSomething isn't working

triageIssue needs to be triaged/prioritizedIssue needs to be triaged/prioritized

Status: Open.

#23247 In run-llama/llama_index;

· laslil7437 opened on Sep 25, 2026

[Feature Request]: Memory (SQLAlchemy chat store) serves edited, replayed and re-owned messages as genuine context; aget() never signals a changed store

enhancementNew feature or requestNew feature or request

triageIssue needs to be triaged/prioritizedIssue needs to be triaged/prioritized

Status: Open.

#23246 In run-llama/llama_index;

· tech4biz-yasha opened on Sep 25, 2026

[Bug]: ReActAgent handoff leaves stale current_reasoning in ctx.store, contaminating the next turn's prompt and memory

Status: Open.

#23243 In run-llama/llama_index;

· BlueX888 opened on Sep 25, 2026

Does a hosted endpoint with nothing to store belong on the OpenAI like page?

Status: Open.

#23236 In run-llama/llama_index;

· TheCheetah11 opened on Sep 25, 2026

create_schema_from_function drops Annotated Field constraints (ge/le/min_length) from tool schema

Status: Open.

#23235 In run-llama/llama_index;

· michaelkillgta opened on Sep 25, 2026

Agent setup_agent permanently writes state_prompt into ChatMemoryBuffer history

Status: Open.

#23234 In run-llama/llama_index;

· michaelkillgta opened on Sep 25, 2026

[Bug]: Nodes split from already-split nodes never get a NEXT relationship, and their start/end_char_idx are relative to the parent chunk

bugSomething isn't workingSomething isn't working

triageIssue needs to be triaged/prioritizedIssue needs to be triaged/prioritized

Status: Open.

#23224 In run-llama/llama_index;

· lucaluo925 opened on Sep 24, 2026

[Bug]: llms-anthropic 0.12.0 rejects claude-opus-5-5 during metadata lookup and agent initialization

Status: Open.

#23222 In run-llama/llama_index;

· pramodc opened on Sep 24, 2026

PairwiseComparisonEvaluator: with enforce_consensus=True, a tie in the original order plus a decisive flipped order returns the flipped judge's score/passing unconverted (verdict inverted)

Status: Open.

#23216 In run-llama/llama_index;

· shaurya416 opened on Sep 24, 2026

[Bug]: Claude Opus 5.5 on Bedrock: BedrockConverse rejects its inference profile IDs, and structured_predict gets 400 on forced toolChoice

Status: Open.

#23207 In run-llama/llama_index;

· kimnamu opened on Sep 23, 2026

[Bug]: TokenBucketRateLimiter hangs forever when one request exceeds the token bucket; SlidingWindowRateLimiter lets it straight through

Status: Open.

#23202 In run-llama/llama_index;

· Harsh23Kashyap opened on Sep 23, 2026

[Bug]: PrevNextNodePostprocessor loses retrieved scores when neighbor windows overlap

Status: Open.

#23199 In run-llama/llama_index;

· HuaTNA opened on Sep 23, 2026