BEZEN Support Chat

    1

    A support chat that reads the customer before it answers: frustrated, giving up or ashamed to ask, what not to do now, when a person should take over, and one human reply.

    Free

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    BEZEN Support Chat

    Example session with this skill installed

    A customer wrote in our chat: "This is the third time I am writing to you about the same double charge." How should we answer?

    • Read your context and instructions
    • Compiled the bezen support chat

    Reading

    • What is happening to them: already said it twice and was not heard. Frustrated.
    • Must not do now: ask again for the order number, or explain how long charges take.
    • Automatic or person: a person. Third contact.
    • Next step: an agent gets back to them today.

    Reply

    "This is already the third time you're writing about the same charge, and it's not OK that it still isn't fixed. I've passed the conversation to an agent, who will get back to you today by 4 pm, without you having to explain from the start."

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Free during launch, until 31 Jan 2027.

    The problem: Repeat contacts, and customers who write in anger.

    What it does: BEZEN connects to your support chat, recognizes a customer who is frustrated, giving up, or ashamed to ask for help, and changes the reply accordingly.

    The result: fewer conversations where the customer gets angrier and angrier until they write in capitals, and fewer cases that reach a human agent when the customer is already angry. Whoever is handed to a person gets there earlier and in a better state.

    Example: A customer writes "This is the third time I'm writing". Instead of another question about the order number, they get an acknowledgment of what happened and an offer to talk to an agent.

    What BEZEN reads in every message, before answering: what is really happening to the customer (frustrated, giving up, ashamed to ask, already said it and wasn't heard), what must not be done now, whether a person is needed, and the next step.

    Before: The customer writes "This is the third time I'm writing to you about the same double charge", and the chat replies: "Hi! To help you, please send your order number and the charge details."

    After: "This is already the third time you're writing about the same charge, and it's not OK that it still isn't fixed. I've passed the conversation to an agent, who will get back to you today by 4 pm, without you having to explain from the start."

    Good for: online stores, service providers, and any business with a chat or WhatsApp.

    Known limitations:

    • It cannot see orders or accounts and never invents a status.
    • It never promises compensation or refunds; what may be offered comes from your rules.
    • It does not replace a human agent; it says when to hand off to one.

    Part of BEZEN (bezenai.com), an emotional-pattern layer for AI agents and bots.

    Cannot access live order databases or account history. Cannot authorize refunds or compensation unless explicitly permitted in your business rules.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 4 days ago

    • Free with an Agensi account

    Permissions

    No special permissions declared or detected

    Listed4 days ago

    What's inside

    Frequently Asked Questions

    Unlike standard bots that only identify ticket categories, this skill performs an internal 'State Analysis' to detect specific emotions like shame or suspicion, ensuring the response addresses the user's feelings rather than just their keywords.

    The skill is framework-agnostic and works with any chat interface, WhatsApp bot, or support ticketing system. You simply input your customer message or transcript into the skill to receive the enhanced analysis and response.

    No, this skill functions as an intelligence layer and does not have access to your private databases. It relies on the conversation transcript and the business rules you provide to generate its responses and escalation triggers.

    You receive the core BEZEN emotional intelligence logic, predefined constraints to prevent repetitive loops, and the specialized 'four-line analysis' framework used to process customer states and escalation needs.

    This skill solves 'loop frustration' by enforcing strict constraints that prevent the agent from asking for information already provided or using empty corporate platitudes when a user is clearly angry.

    Setup is straightforward: you can either pass a single message for a real-time reply or input your current bot instructions to receive a 'BEZEN-enhanced' version of your existing prompt logic.

    New in Agents & Orchestration