Watcher Ops Moderation Reports

Whenever a Watcher takes a meaningful moderation action — issuing a warning, applying a temporary mute, kicking a repeat offender, purging a removed user's messages, or removing a scam link or prohibited content — it posts a structured report into your connected Escalation Group, so your team has visibility without needing to constantly watch every public Community.

Reports are built from templates, not written by a model — every fact (username, warning number, mute duration, purge result, which Watcher acted) is already known before the report is generated, so the wording is predictable and never robotic-sounding filler. Deterministic reports like these never count toward your account's AI Generations usage.

Not every tiny action gets its own report — the goal is a useful operational record, not noise. A report only fires for a genuinely meaningful event: a warning, a mute, a kick, a purge, or a moderation action that couldn't complete due to missing Telegram permissions.

Reports reflect ACTUAL outcomes, never an assumed one — if a kick or a message purge fails (for example, because the Watcher's bot doesn't have the right admin permission in that Community), the report says so honestly instead of claiming success.

Warning and mute history for a member belongs to the Community they're in, not to whichever specific Watcher happened to catch each incident — if a second Watcher in the same Community catches the same person spamming again, it sees the existing history and continues the escalation rather than starting over.

When more than one Watcher is deployed in a Community, only one of them acts on and reports any given violation — you'll never see duplicate warnings, duplicate kicks, or duplicate reports for the same incident.

This is available on the Free Trial too — it's a core part of what makes Watcher Ops useful, not something held back for paid plans.