langflow · Issues· 1071 open

feat: propose an optional Baizhi Agent Toolkit MCP preset

OpenSearch Multi-Model component hardcodes method.name: disk_ann, breaking faiss/nmslib/lucene index creation

SSE mode MCP tools always fail with TimeoutError — HTTP 400 misclassified as transient, blocking SSE fallback

Run Flow component returns empty content when invoked as an Agent tool

[Feature]: docs example for PZERO on OpenAI Compatible provider

Would a high-level architecture diagram be useful for the project documentation?

Serve as A2A - try it

Azure OpenAI Model / Azure OpenAI Embeddings API Request SSL Verify Enable/Disable Feature

TracingService.start_tracers() wraps all 8 tracers in one try/except — one missing optional dependency disables native tracing too

Upgrade Langfuse tracing integration to Python SDK v4

MCP stdio session readiness timeout still hardcoded at 30s with no settings override (configurability gap not addressed by #14161/#14214)

Custom component input field named 'index' (or other reserved Component attribute) silently resolves to the wrong value

Custom component reload silently resolves to the wrong class when a file defines more than one Component subclass

The database flows could not be loaded

Proposal: AnyLLM component — one OpenAI-compatible bundle for 24+ Chinese & international models, with dynamic model discovery