Cron announce 丢消息止血：避免主会话 NO_REPLY 吞掉定时播报

问题/场景：批量 cron announce 注入主会话时，主代理可能返回 NO_REPLY，导致目标频道静默收不到消息。前置条件：使用 `sessionTarget=isolated` + `delivery.mode=announce`，且有可检查的 cron runs/logs。实施步骤：先复现批量运行→对比 run 成功与目标频道无消息→优先启用 direct delivery（`delivery.to` 明确目标）并减少依赖主会话总结→用 bestEffort=false 观测失败。关键点：把“投递成功”与“agent 生成成功”分开验证。验证：run=ok 且目标 topic 实际收到内容。风险：bestEffort=true 会隐藏失败。

GITHUBDiscovered 2026-02-15Author TedwardForma

Prerequisites

Cron jobs are configured with announce delivery and observable target channel/topic.
You can inspect `openclaw cron runs` and gateway logs for each execution.

Steps

Create a controlled batch (2-3 jobs) that fires in the same window to reproduce missing announces.
Compare run records (`status: ok`) with channel delivery reality; mark mismatches as delivery-path failures.
Prefer direct recipient delivery (`delivery.to`) and avoid relying on main-session summarization for critical alerts.
Temporarily set strict failure visibility (avoid best-effort hiding) until delivery path stabilizes.

Commands

openclaw cron runs --id <jobId>

openclaw gateway status

Verify

For at least 3 consecutive runs, each cron run has matching message evidence in the target channel/topic.

Caveats

Issue analysis is based on active bug reports; exact behavior may change after upstream fixes（需验证）.
Do not treat `lastStatus: ok` as delivery success without channel-side evidence.

Source attribution

This tip is aggregated from community/public sources and preserved with attribution.

Open original source ↗

Visit original post