qwen-code SDK v0.1.13 fixes microcompaction to preserve prompt-cache reuse in long agent sessions

github-actions[bot] · ghdev · 2026-09-19

QwenLM released qwen-code TypeScript SDK v0.1.13 (bundling CLI 0.24.1) with two notable fixes: managed auto-memory now respects the memory.enableManagedAutoMemory setting so hosts that disable it no longer get remember/dream requests injected into the system prompt, and size-triggered microcompaction now clears old tool results toward a low watermark instead of rewriting the conversation prefix each turn — preserving provider prompt-cache reuse in long tool-heavy sessions and cutting cached-input costs.

Original post →

More from coding & agent

coding & agent channel →