More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+22 more

    Whatsapp AI Agent

    1

    Generates a working WhatsApp AI agent in under a minute. Bring your own model and memory backend.

    Free

    Security scanned
    Whatsapp AI Agent

    Whatsapp AI Agent

    Example session with this skill installed

    Generate a new WhatsApp agent using local memory. I don't need the productivity tools. Put it in a folder called 'my-support-bot'.

    • Read your context and instructions
    • Compiled the whatsapp ai agent

    I ran the generator for your WhatsApp AI agent in the my-support-bot directory. The project is configured to use Anthropic as the model provider and local SQLite-based memory.

    Project Details

    ComponentSelectionDescription
    Model providerAnthropicUses @ai-sdk/anthropic via the Vercel AI SDK
    Memory backendLocal (SQLite/LibSQL)No account needed, BM25 keyword search, runs fully offline
    Productivity agentOffSkipped — no Gmail/Calendar tools included

    Files generated

    • index.ts — the agent itself: model, memory, routing, and WhatsApp wiring
    • package.json — dependencies scoped to exactly what this build needs: @wappy_ai/core, @wappy_ai/harness, @wappy_ai/whatsapp, @ai-sdk/anthropic, ai, @libsql/client, dotenv
    • .env.sample — only the keys this specific build needs: ANTHROPIC_API_KEY, WHATSAPP_PHONE_NUMBER_ID, WHATSAPP_ACCESS_TOKEN, WHATSAPP_VERIFY_TOKEN, WHATSAPP_APP_SECRET
    • README.md — your project's own next steps
    • WHATSAPP_SETUP.md — the full Meta app + webhook walkthrough

    Next steps

    cd my-support-bot
cp .env.sample .env   # fill in your Anthropic key + WhatsApp credentials
npm install
npm run dev            # boots the server, opens a tunnel, prints the webhook URL for Meta

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

    About this skill

    Here's the rewrite — "scaffold"/"scaffolding" replaced throughout, and "production-ready" softened to something we can actually back up:

    Stop wasting time writing boilerplate for WhatsApp integrations. This skill generates a complete, runnable TypeScript project for building AI agents that live directly on WhatsApp. It uses the open-source Wappy SDK to ensure your data and model keys stay on your own infrastructure rather than routing through third-party middleware.

    What it does

    • Project generation produces a complete, working TypeScript codebase instead of empty templates.
    • Provider selection configures the agent for OpenAI, Anthropic, Gemini, or local-first Ollama.
    • Memory management implements SQLite/LibSQL for local keyword search or Cognee for knowledge graphs.
    • Tool integration sets up optional read-only Gmail and Google Calendar tools for productivity workflows.
    • Environment setup produces a tailored .env.sample containing only the keys required for your specific build.

    How it works

    1. Interactive mode initiates a three-question interview to define the model, memory, and features.
    2. Scripted execution uses flags like --model, --memory, and --productivity for automated CI/CD or non-interactive setups.
    3. Environment configuration prepares the project for local execution via npm run dev with automatic webhook tunneling.

    Frameworks & tools

    TypeScript, Node.js 22+, Wappy SDK, SQLite/LibSQL, Cognee, and the Meta WhatsApp Cloud API.

    Why this beats prompting it yourself

    Prompting an LLM to write a WhatsApp agent often results in outdated API calls or missing webhook logic. This skill provides a verified, structured pipeline that handles message routing and tool-call decisions out of the box.

    Use cases

    • Build a private WhatsApp assistant using local Ollama for maximum privacy.
    • Create a customer support bot with a RAG backend for answering product questions.
    • Develop a personal productivity agent that queries your Google Calendar via chat.

    Known limitations

    Requires Node.js 22 or higher. Users must manually create a Meta Developer App to obtain WhatsApp Cloud API credentials.

    Want this written into AGENSI_SUBMISSION_NOTES.md so it's saved alongside the other listing copy, or is this just for pasting directly into the listing form?

    How to use

    Skills on Agensi run inside your AI agent. Copy the prompt, paste it into your agent, and it connects to Agensi and installs this skill for you. If a payment is needed, your agent shows it to you first.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill passes our trust checks below.

    Security scanned

    Passed our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free forever

    No account required to browse

    Trust & safety

    Security scanned

    Verified clean today

    • Free to download with an account

    Requires

    Node.js

    Listedtoday

    What's inside

    Frequently Asked Questions