fix(llm): strip dangling XML tool call closing tags from text content - #27984

Issue for this PR

Closes #24316

Type of change

• Bug fix
• New feature
• Refactor / code improvement
• Documentation

What does this PR do?

Qwen3 via vLLM/llama.cpp with the hermes tool call parser occasionally appends raw XML tool call closing tags (</parameter>, </function>, </parameter, ␐) as a response termination artifact after natural language text. These leak into the assistant message content and appear as naked XML in the TUI, halting progress.

This adds a stripDanglingXmlArtifacts() helper that strips trailing XML artifacts from streaming text deltas at the step() level in openai-chat.ts before content reaches the consumer. Only trailing artifacts are stripped; mid-text occurrences are preserved.

I understand why this works: the artifacts always appear at the end of a streaming delta chunk, so iterating and stripping trailing matches is sufficient without risking damage to legitimate content.

How did you verify your code works?

• bun test in packages/llm — 16/16 pass
• bun typecheck — clean
• Stress tested with multi-tool-call prompts (5-10 iterations each) against Qwen3/vLLM — no XML leakage observed

Screenshots / recordings

N/A — this is a backend streaming fix.

Checklist

• I have tested my changes locally
• I have not included unrelated changes in this PR