PraisonAI Claw

Connect your AI agents to Telegram, Discord, Slack and more — all from a single command

PraisonAI Claw connects your AI agents to Telegram, Discord, Slack, and more messaging platforms — all from a single command. Install once, launch the dashboard, and add channels from the UI.

Quick Start

1. Install

   pip install "praisonai[claw]"
   
2. Set your API key

   export OPENAI_API_KEY="your-api-key-here"
   
3. Launch

   praisonai claw
   

   Open http://localhost:8082 — the dashboard is live.

Connect a Channel (via UI)

The easiest way to add a messaging channel is through the dashboard:

1. Open the Channels page

   In the dashboard sidebar, click 📡 Channels.

2. Click + Add Channel

   Select a platform from the dropdown:

   Platform	Token needed
   ✈️ Telegram	Bot Token (from @BotFather)
   🎮 Discord	Bot Token (from Developer Portal)
   💬 Slack	Bot Token + App Token
   📱 WhatsApp	Access Token + Phone Number ID
   🔒 Signal	Phone Number + API URL
   💼 Google Chat	Service Account + Space Name
   🟣 Nostr	Private Key + Relay URL

3. Paste your token and click Add

   The dashboard shows setup steps for each platform right in the form. Once added, your bot starts automatically — you’ll see a ● Connected status.

4. Test the connection

   Click 🔍 Test on any channel card to verify your bot is responding. You can also Disable, Restart, or Delete channels from the same card.

Get Platform Tokens

Telegram

1. Open Telegram → search @BotFather
2. Send /newbot → follow the prompts
3. Copy the Bot Token

Discord

1. Go to discord.com/developers → New Application
2. Go to Bot → Reset Token → copy it
3. Enable Message Content Intent under Privileged Gateway Intents
4. Go to OAuth2 → URL Generator → select bot scope → invite to your server

Slack

1. Go to api.slack.com/apps → Create New App → From Scratch
2. OAuth & Permissions → add scopes: chat:write, app_mentions:read
3. Enable Socket Mode → copy the App Token (xapp-...)
4. Install to Workspace → copy the Bot Token (xoxb-...)

CLI Options

praisonai claw                        # Dashboard on port 8082
praisonai claw --port 9000            # Custom port
praisonai claw --host 127.0.0.1      # Localhost only
praisonai claw --app my-dashboard.py  # Custom app file
praisonai claw --reload               # Auto-reload on changes

YAML Agent Mode

For a quick agent without any Python:

# agents.yaml
name: Research Assistant
instructions: |
  You are a helpful research assistant.

model: gpt-4o-mini

welcome: "👋 Hello! How can I help?"

starters:
  - label: "Explain a concept"
    message: "Explain machine learning vs deep learning"
    icon: "🧠"
  - label: "Search the web"
    message: "Search for the latest AI news"
    icon: "🔍"

tools:
  - web_search
  - calculate

features: true
datastore: json

aiui run agents.yaml

YAML Reference

Key	Type	Description
name	string	Agent display name
instructions	string	System prompt
model	string	LLM model (e.g., gpt-4o-mini)
welcome	string	Message shown on connect
goodbye	string	Message shown on disconnect
starters	list	Conversation starters (label, message, icon)
profiles	list	Selectable personas (name, description, icon)
tools	list	Built-in tools: web_search, calculate, etc.
features	boolean	Enable all protocol features (memory, sessions, schedules, guardrails, skills, approvals, hooks, workflows)
datastore	string	Persistence: json, memory, or json:/path

Connect Channels via Code

For advanced use, connect channels directly in Python:

Telegram

import os
from praisonaiagents import Agent
from praisonai.bots.telegram import TelegramBot

agent = Agent(
    name="Assistant",
    instructions="You are a helpful assistant.",
    llm="gpt-4o-mini",
)

bot = TelegramBot(
    agent=agent,
    token=os.environ["TELEGRAM_BOT_TOKEN"],
)
bot.start()

import os
from praisonaiagents import Agent
from praisonai.bots.discord import DiscordBot

agent = Agent(
    name="Assistant",
    instructions="You are a helpful assistant.",
    llm="gpt-4o-mini",
)

bot = DiscordBot(
    agent=agent,
    token=os.environ["DISCORD_BOT_TOKEN"],
)
bot.start()

import os
from praisonaiagents import Agent
from praisonai.bots.slack import SlackBot

agent = Agent(
    name="Assistant",
    instructions="You are a helpful assistant.",
    llm="gpt-4o-mini",
)

bot = SlackBot(
    agent=agent,
    token=os.environ["SLACK_BOT_TOKEN"],
)
bot.start()

Docker

# Dashboard
docker run -p 8082:8082 -e OPENAI_API_KEY=sk-xxx praisonai:claw

# Telegram bot
docker run -e OPENAI_API_KEY=sk-xxx -e TELEGRAM_BOT_TOKEN=... \
  praisonai:claw praisonai bot telegram

# Discord bot
docker run -e OPENAI_API_KEY=sk-xxx -e DISCORD_BOT_TOKEN=... \
  praisonai:claw praisonai bot discord

Dashboard Pages

The praisonai claw dashboard comes with 13 built-in pages:

Page	What it does
💬 Chat	AI agent chat with streaming
📡 Channels	Connect and manage messaging platforms
🤖 Agents	Create and manage AI agents
⚡ Skills	Browse tools and plugins
🧠 Memory	View and manage agent memory
📚 Knowledge	Knowledge base and RAG
🛡️ Guardrails	Input/output safety rules
⏰ Cron	Scheduled agent jobs
📋 Sessions	Conversation history
📈 Usage	Token usage and costs
⚙️ Config	Runtime configuration
📜 Logs	Server logs
🐛 Debug	System debug info

What’s Included

pip install "praisonai[claw]" installs:

Component	What you get
Dashboard	Full web UI with 13+ admin pages
Bot Channels	Telegram, Discord, Slack, WhatsApp, Signal, Google Chat, Nostr
Agent Runtime	Multi-agent orchestration with 100+ tools
Gateway	API server with SSE streaming
Web Search	Tavily, Crawl4AI, DuckDuckGo
Session Management	Per-user conversation isolation

Related

• Bot Operating System — bot architecture and lifecycle
• AgentOS — operating system for AI agents