Qwen 3.5 template bug breaks KV-cache reuse

The "thinking" model era has introduced a new class of brittle failure modes where performance is dictated by Jinja template serialization discipline rather than raw model weights.

• –Empty `<think>` tags in history create non-identical prompt prefixes, resulting in near-zero cache hit rates and "Time To First Token" (TTFT) spikes.
• –Agent workflows are disproportionately affected because frequent tool calls and follow-ups compound the cumulative impact of template-induced drift.
• –This issue highlights the fragility of the "KV-cache contract" in reasoning models, where even minor formatting inconsistencies can wipe out the benefits of prefix caching.
• –Developers using Qwen 3.5 locally should prioritize patching their `chat_template.jinja` before attempting more complex infrastructure-level cache workarounds.