Canonical: https://botbento.com/usecases/route-incoming-channel-messages-to-the-right/
Format: Markdown representation of the public HTML page.

# Route incoming channel messages to the right workflow

[Operations](https://botbento.com/usecases/ops/) · Intermediate

Watches incoming messages and drafts the next action for approval

An agent keeps an eye on several incoming message channels and sorts each new item: a question worth answering, a task worth filing, or noise to leave alone. It remembers how similar items were routed before so repeat cases get the same treatment, and it prepares the reply or task without sending anything on its own.

## Quick answers

**What does it do?** Watches incoming messages and drafts the next action for approval. Result: Routing entry with approval request.

**What does it need?** Tools: Slack, Telegram, Email, Notes. Intermediate: some setup.

**How do I run it?** When something happens: New message in a watched channel. Copy the prompt below into a bot in BotBento.

## How it works

1. When something happens: New message in a watched channel
2. Then: Classify intent and draft next action
3. Result: Routing entry with approval request

## Prompt to start with

```text
Watch the message channels listed in your notes and treat every new item as a routing decision. Classify each one as a question, a task, or noise, and write a one-line reason for the choice. Keep a short memory file of prior decisions and consult it before classifying so similar items land in the same place. Draft the reply or the task text, file the routing entry, and then stop. Do not send, post, pay, delete, or contact anyone until the user approves that specific item; queue those actions and show the draft first.
```

## Tools it uses

Slack, Email, Telegram, Notes

## Works with

BotBento

## Tags

triage, routing, event-driven, approvals, memory

## Source

[Ask HN: What's your current AI agentic setup?](https://news.ycombinator.com/item?id=50366117) — Hacker News · checked 2026-11-04

Summarized in our own words. Any results described there are the authors’ own, not measured by BotBento.

Verified 2026-11-04: passed every automated gate and a second-model review (kimi-code/k3.1-pro). Automated, not human review. [How verification works](https://botbento.com/usecases/verification/)

By BotBento Editorial · Added 2026-11-04 · Updated 2026-11-04 · [Suggest a correction](https://botbento.com/contact/)

## Related use cases

- [Automated background task executor with supervised approvals](https://botbento.com/usecases/automated-background-task-executor-with-supervised-approvals/) — Keep recurring agent jobs running on schedules, with risky fixes gated by approval
- [Handling internal operational requests](https://botbento.com/usecases/handling-internal-operational-requests/) — Intake employee inquiries, check operational records, and prepare approvals for review
- [Multi-agent task dispatcher and coordinator](https://botbento.com/usecases/multi-agent-task-dispatcher-and-coordinator/) — Break down compound user requests and route subtasks across specialized agent threads
