AI AGENT TEMPLATE
An AI agent that watches tomorrow's weather and alerts you if a scheduled bike tour is at risk, with draft customer communication ready to send.
Bike tour weather decisions are higher-stakes than most operators treat them. A guided city bike tour in heavy rain isn't impossible, but it's miserable, and the reviews that follow will punish you for running it anyway. Meanwhile, cancelling late — after customers have already paid, packed lunches, and got on the subway to your meeting point — generates its own kind of negative review.
The Bike Tour Weather Alert Agent gets the decision made early and the communication out fast. The agent watches the forecast continuously; when rain probability, wind speed, heat index, or UV exceeds your per-tour thresholds, it drafts customer-specific emails offering a reschedule to the next available equivalent tour, with refund as a fallback option only.
The reschedule-first framing recovers 60-80% of affected revenue versus 15-25% for refund-first policies. For bike tour operators, where unit economics are tight and marketing acquisition is expensive, that revenue recovery is the difference between a profitable weather-volatile month and a painful one. The agent also segments by tour type — an indoor-heavy e-bike tour tolerates weather that shuts down a long countryside ride.
Below is the full agent spec: trigger thresholds, tools, message templates, and how to configure per-tour-type weather criteria and reschedule logic.
"Every 4pm, pull tomorrow's forecast for each tour location. If rain probability >70% or conditions dangerous (storms, extreme heat), flag the affected tours and draft a customer-communication email offering reschedule or refund."
Paste this into Dash Agents. Dash reads the prompt, picks the right tools, assembles the logic, and creates a ready-to-run agent in seconds.
The agent uses the standard Dash Agents tool library. Every tool call is logged.
On trigger, the agent runs these steps in order. Any step can be customised by re-prompting.
Get list of scheduled tours for tomorrow with customer contact info.
Forecast for each tour location: rain %, wind, temperature, storm alerts.
Safe / marginal / unsafe. Thresholds configurable.
For affected tours, draft personalised email offering reschedule or refund.
Send full report + draft email for approval.
Execute the outreach, log on each booking.
Customisations by re-prompting in plain English: change the trigger timing or conditions, adjust the recipient list, tune thresholds, modify tone, add approval gates, connect additional channels (SMS, WhatsApp) if available. Run in approval mode for the first week to build confidence before switching to autonomous.
Structured automation beats manual follow-up in three ways:
Set this agent up on your first week. Run in approval mode for five days while you watch the outputs, then flip to autonomous. The ROI compounds across a full season — a quiet automation that simply does the right thing, every time, without anyone having to remember.
Decide the day before, not the morning of. Watch the forecast at a consistent time (afternoon before the tour). If weather is clearly unsafe, contact customers that evening with reschedule or refund options. Marginal conditions: contact customers with what to expect and equipment recommendations. This approach reduces last-minute scrambles and builds customer trust.
Dash Agents handle the repetitive work so your team can focus on customers. Start your free trial and build your first agent in minutes.
