[Feature] Let sub-agents resume from checkpoint after quota exhaustion instead of restarting

提交前确认 · Pre-submission checklist

• 我已搜索过现有 issue，确认这不是重复提议 / I searched existing issues and confirmed this isn't a duplicate.
• 我已阅读 CONTRIBUTING.md / I've read CONTRIBUTING.md.

问题类别 · Category

对话 / Agent 交互 · Agent chat / interaction

涉及的 Agent 框架 · Agent framework

ZCode Agent（自研）

使用场景 · Use case

When a sub-agent runs out of quota mid-task, resuming after the 5-hour window doesn't continue from where it left off — most of the context is lost. I have to say "continue" and it re-runs a large chunk of the task from scratch, burning tens of millions of tokens redoing work that was already done or inferring on the half work it had prior, would be nicer to just resume it like the chat does per subagent

建议方案 · Proposal

Persist sub-agent state/context the same way a main chat session is persisted, so when quota resets, the sub-agent can resume exactly where it stopped instead of restarting.

预期价值 · Expected value

Avoids wasting large amounts of tokens re-running already-completed sub-agent work after every quota reset.

你认为的优先级 · Your perceived priority

None

你使用的 ZCode 版本 / 环境 · ZCode version / environment

(the newest as of august 11th 2026, at 3:10 am EST exactly.. sorry idk exactly)

补充材料 · Additional context

No response

No implementation files, tests, or entry points are identified in the issue. Start by reading the proposal and CONTRIBUTING.md, then locate how sub-agent state and main chat sessions are persisted; done means a quota-exhausted sub-agent resumes its prior context without rerunning completed work.