ibrahimhajjaj

Wa Agent — AI skill for Claude Code

AI community

Framework for building autonomous AI agents on WhatsApp.

How to install Wa Agent

This entry records only its repository, not the path inside it, so there is no exact command to give. Open ibrahimhajjaj/wa-agent and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Wa Agent does

Framework for building autonomous AI agents on WhatsApp.

Alternatives in AI

WeKnora — Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, an 26.9k ★
Ralph For Claude Code — by Frank Bria - An autonomous AI development framework that enables Claude Code to work iteratively on project 8k ★
Osaurus — Own your AI. The native macOS harness for AI agents -- any model, persistent memory, autonomous execution, cry 5.1k ★

README

wa-agent

Build autonomous AI agents that live on WhatsApp.

wa-agent gives you the plumbing so you can focus on the agent. You write a YAML file describing your agent's personality, tools, and routing rules — wa-agent handles the WhatsApp connection, message queuing, conversation memory, tool execution, and everything else.

Powered by [Vercel AI SDK v6](https://sdk.vercel.ai/) for the agent layer and [wu-cli](https://github.com/ibrahimhajjaj/wu-cli) for battle-tested WhatsApp internals.

Quick Start

npx wa-agent init my-bot
cd my-bot
npm install

Edit `wa-agent.yaml` with your API keys:

version: 1

agents:
  dir: ./agents

webSearch:
  provider: tavily
  apiKey: "${TAVILY_API_KEY}"

fetchUrl:
  provider: jina        # jina (default) | local
  # apiKey: "${JINA_API_KEY}"  # optional — works without key (20 RPM), with key gets 500 RPM

Edit `agents/my-agent.yaml`:

name: my-agent
description: My WhatsApp AI agent

llm:
  provider: anthropic
  model: claude-sonnet-4-20250514
  temperature: 0.7

personality: |
  You are a helpful WhatsApp assistant. Be concise — this is WhatsApp,
  not email. Use short paragraphs and get to the point quickly.

tools:
  - send-message
  - web-search
  - fetch-url

routing:
  - type: default
    match: "*"

memory:
  conversationWindow: 20
  summarizeAfter: 100
  userProfiles: true

Start it up:

npx wa-agent start

First run will show a QR code — scan it with WhatsApp to link the agent. wu-cli handles reconnection automatically after that. You only need to re-scan if WhatsApp terminates the linked device (rare, usually from inactivity or logging out manually).

Why wa-agent?

Every WhatsApp bot repo out there is the same thing: glue code piping messages to OpenAI. No memory across conversations. No tool use. No multi-agent routing. No rate limiting. No way to handle a busy group without burning through your API budget.

wa-agent is a framework, not a chatbot template. It gives y