qwen-code PR closes recovery gap: rewritten commands need fresh approval, replay from durable receipts
wenshao · ghdev · 2026-11-02
PR #16116 on QwenLM/qwen-code completes the A3 half of the G3 Step 3 recovery matrix for managed command execution: a Hook-rewritten command input is treated as a new, independently approved step rather than inheriting a stale Allow; completed Hooks and command steps replay from durable results after restart without physical re-execution; and a skill expanded into two command steps survives a crash while preserving the second step's identity and deadline. It ships a packaged failover test arm, durable evidence collection, regression witnesses, and bilingual design docs, addressing safety-critical gaps tracked in issue #14481.
More from coding & agent
- From chatbot to agent: how tools, planning and memory were added step by step — _jaydeepkarale · 2026-11-04
- Binance's $7B hack reportedly started with a tampered Homebrew install during vibe coding — xuanalogue · 2026-11-04
- $100 of GPU time edits model weights to cut thinking tokens by 40% — jarrodwatts · 2026-11-04
- Cua Driver integrates with Google Antigravity for background cross-app computer agents — m4rkmc · 2026-11-04
- Sentdex: OpenHands was the first reasonably useful agent to abstract away the code editor — rajistics · 2026-11-04
- Security advisory: run AI agents in managed cloud environments, not employee endpoints — amasad · 2026-11-04