# Channels Adragent meets you where you already work — chat inside the console, email, WhatsApp, SMS. You decide which to use moment by moment; Adragent prepares the same kind of operation regardless of where the request starts. ## Chat *** `In-console · daily driver` Chat lives in the console next to the rest of your work. You ask, Adragent answers, you stay in flow. It remembers what you have been talking about, so the next sentence builds on the last. * *"Show me Hotel Sunrise rate plans below 80 EUR for May."* * *"Find every property in Antalya that is missing a hero image."* * *"Draft a 15% weekend discount for the EU member rate plans."* * *"Now do the same for Hotel Sea Breeze."* What chat is best at: * **Bulk operations.** *"Stop-sale rooms 12, 14, 16 from June 1 to 8"* — 3 rooms × 8 nights = 24 inventory cells set to zero with one preview, one *Confirm*. * **Multi-step workflows.** *"Find properties without images and queue them for content review"* — find, filter, batch action, in one go. * **Quick exploration.** Shaped data instead of a screen to interpret. * **Cross-domain orchestration.** A pricing change, a distribution scope and an audience filter, expressed as a single intent. *** ## Email *** `Intent in only` Forward what already arrived in your inbox. Adragent reads it, extracts the structured intent, and prepares the operation. Replies happen in the console, not back to the email thread. * **Forward a promo email** to your organization's Adragent inbox. Adragent picks out the rate, dates and properties, and drafts the promotion. * **Forward a contract** for a new supplier. Adragent extracts parties, allotments and markup terms, and prepares an onboarding draft. * **Paste an email body** directly into chat instead of forwarding — same outcome. {% hint style="info" %} **Email is one-way.** You forward in; Adragent prepares the operation; you confirm in the console (or chat / WhatsApp / SMS — whichever is at hand). Adragent never replies to the original email thread. {% endhint %} *** ## WhatsApp *** `Off-desk` When you are not at your desk, WhatsApp keeps you moving. Your WhatsApp number is bound once. From then on, WhatsApp behaves like chat — same tools, same preview and approval pattern. * *"Stop-sale Hotel Sunrise tonight."* * *"Any failed channel pushes in the last hour?"* * *"Approve"* / *"Reject"* — for an operation Adragent staged earlier in chat. *** ## SMS *** `Approval-only · narrowest channel` SMS is intentionally narrow. Two patterns: * **Outbound alerts.** A payment chargeback opens. A rate distribution fails. A contract expires next week. Adragent reaches you wherever you are. * **Inbound approvals.** When an operation needs your green light and you are not in the console, an SMS with a short reply token (`YES ` / `NO `) lets you approve from any phone. {% hint style="warning" %} **Approval codes are scoped, single-use and time-limited.** A code lets you approve exactly one specific operation Adragent staged for you. You cannot reuse a code, and it expires after a short window. {% endhint %} *** ## What stays the same across channels * **It is still your organization's data.** Adragent never reaches across organization boundaries. * **It is still the same workflow.** Requests, previews, approvals, and results appear in the same product views. * **It is still clear.** Adragent responds in the language used for the request, including the confirmation prompt. ## Where to next * **Why every change waits for your *****yes*** → [Confirm-before-apply](/console/adragent/confirm-before-apply.md) * **What Adragent can actually do** → [Tools](/console/adragent/tools.md) --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://adrasis.gitbook.io/console/adragent/channels.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.