About the People.ai Automation Library

This library contains ready-to-import workflow automations that leverage the People.ai MCP (Model Context Protocol) integration alongside AI models, CRM data, and communication tools to deliver actionable sales intelligence.

Each workflow is a complete automation that can be imported into your preferred platform. Workflows combine People.ai's engagement data with Claude AI to generate insights like daily sales digests, meeting briefs, pipeline coaching reports, and account risk alerts.

Supported Platforms:
Any (JSON import) Workato — Coming Soon Zapier — Coming Soon Others — Coming Soon

How it works:
1. Browse the library and pick a workflow
2. Download the workflow JSON
3. Import into your automation platform
4. Connect your credentials (People.ai MCP, Anthropic, Slack, etc.)
5. Configure schedule and channel settings
6. Activate the workflow

Prerequisites:
• An automation platform instance (self-hosted or Cloud)
• People.ai MCP credentials
• Anthropic API key for Claude AI
• Slack Bot token (for most workflows)
• Additional credentials as noted per workflow

Maintained by: AI Innovation Team
Slack: #automation-workflows

People.ai MCP Setup Guide

The People.ai Model Context Protocol (MCP) server is the core data source for every workflow in this library. It provides AI agents with direct access to account activity, engagement data, opportunity status, and contact intelligence.

What You Can Do

• Automate sales intelligence workflows like daily account health checks, weekly opportunity reviews, or risk alerts
• Build AI-powered chatbots for Slack or Teams that answer questions with real CRM data
• Create coaching automation that analyzes opportunities and generates weekly insights
• Generate automated reports combining People.ai insights with data from other systems
• Trigger actions based on sales data like sending alerts when accounts show low engagement

Available MCP Tools

The People.ai MCP server exposes these tools to connected agents:

1. Obtain MCP Credentials

1. Contact your People.ai administrator or CSM to request MCP API access
2. You'll receive two credentials:
   • Client ID — your organization's MCP client identifier (e.g., Yl1JnBI...)
   • Client Secret — authentication secret (e.g., 2rT0SWrgR...)
3. Store these credentials securely — you'll need them for configuration

2. Configure in n8n

n8n v1.19.0+ has a built-in MCP Client tool. Here's the exact configuration:

1. Create a new workflow or open an existing one
2. Add an AI Agent node (Claude is recommended for complex sales analysis; OpenAI GPT-4 also works)
3. In the AI Agent node, click Add Tool and select MCP Client
4. Configure the MCP Client:

   Connection Name	People.ai MCP Multi-Header
   Server URL	https://mcp.people.ai/mcp
   Authentication	Multiple Headers Auth

5. Configure three authentication headers:

   Header	Name	Value
   Header 1	PAI-Client-Id	[Your Client ID]
   Header 2	PAI-Client-Secret	[Your Client Secret]
   Header 3	Content-Type	application/json

6. Under Allowed HTTP Request Domains, select All or add mcp.people.ai
7. Click Save to create the credential

3. Configure Your AI Agent

Add a system prompt to your AI Agent that instructs it how to use People.ai tools:

You are a sales intelligence assistant with access to People.ai CRM data.
When users ask about accounts, opportunities, or customer activity:
1) Use People.ai MCP tools to search and retrieve data
2) Analyze the information and provide strategic insights
3) Highlight risks, next steps, and key topics
4) Format responses clearly and actionably

4. Other Platforms

LangGraph (Python)

pip install langchain-mcp-adapters

from langchain_mcp_adapters.client import MultiServerMCPClient

client = MultiServerMCPClient({
    "peopleai": {
        "url": "https://mcp.people.ai/mcp",
        "transport": "sse",
        "headers": {
            "PAI-Client-Id": os.environ["PEOPLEAI_CLIENT_ID"],
            "PAI-Client-Secret": os.environ["PEOPLEAI_CLIENT_SECRET"],
            "Content-Type": "application/json",
        }
    }
})
tools = await client.get_tools()
# Bind tools to your LLM: llm.bind_tools(tools)

Claude Agent SDK (Python)

pip install claude-agent-sdk anthropic

from claude_agent_sdk import Agent
import anthropic

agent = Agent(
    client=anthropic.Anthropic(),
    model="claude-sonnet-4-20250514",
    system="You are a sales intelligence assistant...",
    tools=[...],  # local tools
    mcp_servers=[{
        "name": "peopleai",
        "url": "https://mcp.people.ai/mcp",
        "headers": {
            "PAI-Client-Id": os.environ["PEOPLEAI_CLIENT_ID"],
            "PAI-Client-Secret": os.environ["PEOPLEAI_CLIENT_SECRET"],
            "Content-Type": "application/json",
        }
    }],
)

OpenAI Agents SDK (Python)

pip install openai-agents

from agents import Agent
from agents.mcp import MCPServerSse

peopleai_mcp = MCPServerSse(
    name="peopleai",
    url="https://mcp.people.ai/mcp",
    headers={
        "PAI-Client-Id": os.environ["PEOPLEAI_CLIENT_ID"],
        "PAI-Client-Secret": os.environ["PEOPLEAI_CLIENT_SECRET"],
        "Content-Type": "application/json",
    },
)

agent = Agent(
    name="Sales Intelligence",
    instructions="You are a sales intelligence assistant...",
    tools=[...],  # local tools
    mcp_servers=[peopleai_mcp],
)

Make / Power Automate / Zapier

5. Environment Variables

Set these for Python-based workflows (LangGraph, Claude Agent, OpenAI Agent):

export PEOPLEAI_MCP_URL="https://mcp.people.ai/mcp"
export PEOPLEAI_CLIENT_ID="your-client-id"
export PEOPLEAI_CLIENT_SECRET="your-client-secret"

# Also needed for the LLM:
export ANTHROPIC_API_KEY="sk-ant-..."   # for Claude Agent / LangGraph (recommended)
export OPENAI_API_KEY="sk-..."          # for OpenAI Agent

6. Test Your Connection

Try these example queries in n8n or your agent to verify the connection:

• "Find the account for [Company Name]"
• "What are the risks for opportunity [ID]?"
• "Show me recent activity for [Account Name]"

Check the execution log to verify: MCP tools appear in available tools, AI Agent calls them successfully, and data is returned.

7. Troubleshooting

8. Combining MCP with REST API

For best results, use the People.ai Insights Export REST API for bulk data extraction and MCP tools for deep intelligence on specific items. Example pattern:

1. REST API pulls all opportunities closing this quarter
2. n8n logic calculates risk scores
3. MCP tools analyze the top 10 high-risk opportunities
4. n8n formats and sends the report

This hybrid approach combines the efficiency of bulk APIs with the intelligence of AI-powered analysis.

Slack Bot Setup Guide

Most workflows in this library deliver results via Slack. Follow these steps to create and configure a Slack bot with the right permissions.

1. Create a Slack App

1. Go to api.slack.com/apps
2. Click Create New App → From scratch
3. Name it (e.g., Automation Catalog Bot) and select your workspace
4. Click Create App

2. Configure Bot Token Scopes

Navigate to OAuth & Permissions in the sidebar, scroll to Bot Token Scopes, and add:

3. Install to Workspace

1. Scroll up on the OAuth & Permissions page
2. Click Install to Workspace
3. Review the permissions and click Allow
4. Copy the Bot User OAuth Token (starts with xoxb-)

4. Invite Bot to Channels

The bot can only post to channels it's been invited to:

• Open the target Slack channel
• Type /invite @Automation Catalog Bot (or your bot's name)
• For DM-based workflows, no invite is needed — the bot opens DMs directly using im:write

5. Configure in Your Platform

n8n

1. Go to Settings → Credentials → Add Credential
2. Select Slack API
3. Paste your xoxb- token in the Access Token field
4. Click Save — the credential is now available to all Slack nodes

LangGraph / Claude Agent / OpenAI Agent (Python)

export SLACK_BOT_TOKEN="xoxb-your-token-here"

Make / Power Automate / Zapier

• Make: Slack connection → paste Bot Token
• Power Automate: Use the Slack connector with OAuth or paste token in HTTP action
• Zapier: Connect Slack account via OAuth flow (scopes configured in the app)

6. Bot Display Name & Icon

To customize how the bot appears in Slack:

1. Go to your app at api.slack.com/apps
2. Click Basic Information in the sidebar
3. Scroll to Display Information
4. Set the app name, description, icon, and background color
5. Each workflow in this library uses a unique bot name (Aria, Watchdog, Prospector, etc.) via the chat:write.customize scope

7. Troubleshooting

Microsoft Teams Setup Guide

Workflows in this library can deliver results to Microsoft Teams via two methods: Incoming Webhooks (quick setup, channel-only) or a Bot Framework bot (full-featured, supports DMs and adaptive cards).

Option A: Incoming Webhooks

Fastest way to post messages to a Teams channel. No Azure registration needed.

1. Create a Webhook in Teams

1. Open the target Teams channel
2. Click ⋮ (more options) next to the channel name → Connectors (or Manage channel → Connectors)
3. Search for Incoming Webhook and click Configure
4. Name it (e.g., Automation Catalog) and optionally upload an icon
5. Click Create
6. Copy the Webhook URL — it looks like:
   https://outlook.office.com/webhook/...

2. Send Messages via Webhook

POST a JSON payload to the webhook URL. Teams supports two formats:

Simple Text (MessageCard)

POST {webhook_url}
Content-Type: application/json

{
  "@type": "MessageCard",
  "summary": "Silence Alert",
  "themeColor": "FF6B6B",
  "title": "🔴 SILENCE ALERT — 3 accounts need attention",
  "text": "**Globex Industries** — 18 days silent..."
}

Adaptive Card (richer formatting)

POST {webhook_url}
Content-Type: application/json

{
  "type": "message",
  "attachments": [{
    "contentType": "application/vnd.microsoft.card.adaptive",
    "content": {
      "$schema": "http://adaptivecards.io/schemas/adaptive-card.json",
      "type": "AdaptiveCard",
      "version": "1.4",
      "body": [
        { "type": "TextBlock", "text": "Sales Digest", "weight": "Bolder", "size": "Large" },
        { "type": "TextBlock", "text": "Your daily summary...", "wrap": true }
      ]
    }
  }]
}

3. Configure in Your Platform

n8n

1. Add an HTTP Request node
2. Method: POST
3. URL: paste your webhook URL
4. Body: JSON with MessageCard or Adaptive Card payload

Python (LangGraph / Claude Agent / OpenAI Agent)

import os, json
from urllib.request import Request, urlopen

webhook_url = os.environ["TEAMS_WEBHOOK_URL"]
payload = json.dumps({
    "@type": "MessageCard",
    "summary": "Alert",
    "title": "🔴 Silence Alert",
    "text": message_content,
}).encode()

req = Request(webhook_url, data=payload,
              headers={"Content-Type": "application/json"})
urlopen(req)

Make / Power Automate / Zapier

• Make: HTTP module → POST to webhook URL with JSON body
• Power Automate: Use the built-in "Post message in a chat or channel" action (no webhook needed — native Teams connector)
• Zapier: Webhooks by Zapier → POST to webhook URL, or use Microsoft Teams integration

Option B: Bot Framework (Azure AD)

Full-featured bot with DM support, adaptive cards, and proactive messaging. Required for workflows that deliver personalized content to individual users.

1. Register an Azure AD App

1. Go to Azure Portal → App Registrations
2. Click New registration
3. Name: Automation Catalog Bot
4. Supported account types: Single tenant (your org only)
5. Click Register
6. Note the Application (client) ID and Directory (tenant) ID
7. Go to Certificates & secrets → New client secret — copy the secret value

2. API Permissions

In your app registration, go to API permissions → Add a permission → Microsoft Graph:

Click Grant admin consent after adding permissions (requires admin role).

3. Create a Bot Channel Registration

1. In Azure Portal, search for Azure Bot and click Create
2. Link it to the app registration you created in Step 1
3. Under Channels, enable Microsoft Teams
4. Under Configuration, set the messaging endpoint if using a custom bot service

4. Install Bot in Teams

1. Go to Teams Developer Portal
2. Create a new app or import the bot
3. Under App features → Bot, link your Azure Bot registration
4. Publish to your org or sideload for testing
5. Users can then find and chat with the bot, or add it to channels

5. Configure in Your Platform

n8n

1. Add a Microsoft Teams node
2. Create a credential with your Client ID, Client Secret, and Tenant ID
3. Select action: Send Message to channel or chat

Python (LangGraph / Claude Agent / OpenAI Agent)

export TEAMS_TENANT_ID="your-tenant-id"
export TEAMS_CLIENT_ID="your-client-id"
export TEAMS_CLIENT_SECRET="your-client-secret"

# Python example — send via Graph API
import msal, requests

app = msal.ConfidentialClientApplication(
    os.environ["TEAMS_CLIENT_ID"],
    authority=f"https://login.microsoftonline.com/{os.environ['TEAMS_TENANT_ID']}",
    client_credential=os.environ["TEAMS_CLIENT_SECRET"],
)
token = app.acquire_token_for_client(
    scopes=["https://graph.microsoft.com/.default"]
)

requests.post(
    f"https://graph.microsoft.com/v1.0/teams/{team_id}/channels/{channel_id}/messages",
    headers={"Authorization": f"Bearer {token['access_token']}"},
    json={"body": {"content": message_content}},
)

Power Automate (native)

• Power Automate has a first-party Teams connector — no bot registration needed
• Use "Post message in a chat or channel" action
• Supports adaptive cards via "Post adaptive card in a chat or channel"
• For DMs, use "Post message in a chat or channel" with Flow bot as the poster

6. Troubleshooting

Google Chat Setup Guide

Workflows can deliver results to Google Chat via Incoming Webhooks (quick, space-only) or a Google Chat App (DM support, cards, slash commands).

Option A: Incoming Webhooks

1. Create a Webhook in Google Chat

1. Open the Google Chat space you want to post to
2. Click the space name at the top → Apps & integrations
3. Click + Add webhooks
4. Name it (e.g., Automation Catalog) and optionally set an avatar URL
5. Click Save
6. Copy the Webhook URL — it looks like:
   https://chat.googleapis.com/v1/spaces/.../messages?key=...&token=...

2. Send Messages via Webhook

Simple Text

POST {webhook_url}
Content-Type: application/json

{
  "text": "🔴 *SILENCE ALERT* — 3 accounts need attention\n\n*Globex Industries* — 18 days silent..."
}

Card Message (richer formatting)

POST {webhook_url}
Content-Type: application/json

{
  "cardsV2": [{
    "cardId": "silence-alert",
    "card": {
      "header": {
        "title": "Silence Alert",
        "subtitle": "3 accounts need attention",
        "imageUrl": "https://example.com/icon.png"
      },
      "sections": [{
        "header": "🔴 Critical",
        "widgets": [{
          "decoratedText": {
            "topLabel": "Globex Industries — $340,000",
            "text": "18 days silent (normally 3-day cadence)"
          }
        }]
      }]
    }
  }]
}

3. Platform Configuration

n8n

1. Add an HTTP Request node
2. Method: POST, URL: your webhook URL
3. Body: JSON with text or cardsV2 payload

Python (LangGraph / Claude Agent / OpenAI Agent)

import os, json
from urllib.request import Request, urlopen

webhook_url = os.environ["GOOGLE_CHAT_WEBHOOK_URL"]
payload = json.dumps({"text": message_content}).encode()

req = Request(webhook_url, data=payload,
              headers={"Content-Type": "application/json"})
urlopen(req)

Make / Power Automate / Zapier

• Make: HTTP module → POST to webhook URL
• Power Automate: HTTP action → POST with JSON body
• Zapier: Webhooks by Zapier → POST, or Google Chat integration

Option B: Google Chat App

Full-featured app with DM support, interactive cards, and slash commands. Required for personalized delivery to individual users.

1. Create a Google Cloud Project

1. Go to Google Cloud Console → Create Project
2. Name it (e.g., automation-catalog-bot)
3. Enable the Google Chat API: search in APIs & Services → Library

2. Configure the Chat App

1. Go to Google Chat API → Configuration
2. Set the App name and Avatar URL
3. Under Functionality, enable:
   • Receive 1:1 messages
   • Join spaces and group conversations
4. Under Connection settings, choose App URL (for webhook-driven) or Cloud Pub/Sub
5. Set Visibility to your organization

3. Service Account for Proactive Messaging

To send DMs without user interaction (proactive messaging), you need a service account:

1. Go to IAM & Admin → Service Accounts → Create
2. Name it chat-bot-sender
3. Create a JSON key and download it
4. Enable domain-wide delegation in Google Workspace Admin if needed

4. API Scopes

5. Sending DMs via Chat API

from google.oauth2 import service_account
from googleapiclient.discovery import build

creds = service_account.Credentials.from_service_account_file(
    os.environ["GOOGLE_CHAT_SA_KEY"],
    scopes=["https://www.googleapis.com/auth/chat.bot"],
)
chat = build("chat", "v1", credentials=creds)

# Create or find DM space with user
dm = chat.spaces().setup(body={
    "space": {"spaceType": "DIRECT_MESSAGE"},
    "memberships": [{"member": {"name": f"users/{user_email}", "type": "HUMAN"}}],
}).execute()

# Send message
chat.spaces().messages().create(
    parent=dm["name"],
    body={"text": digest_content},
).execute()

6. Troubleshooting

Email Delivery Setup Guide

Email is the universal fallback — every workflow can deliver via SMTP. This guide covers Gmail, Outlook/M365, SendGrid, and generic SMTP.

Option A: Gmail / Google Workspace

1. Create an App Password

Google blocks "less secure app" logins. Use an App Password instead:

1. Go to myaccount.google.com/security
2. Ensure 2-Step Verification is enabled
3. Go to App Passwords
4. Select app: Mail, device: Other (name it Automation Catalog)
5. Click Generate — copy the 16-character password

SMTP Settings

Alternative: Gmail API (OAuth)

For Google Workspace orgs, the Gmail API with a service account avoids App Passwords:

1. Enable the Gmail API in Google Cloud Console
2. Create a service account with domain-wide delegation
3. Grant the https://www.googleapis.com/auth/gmail.send scope
4. Use the service account to send on behalf of users

Option B: Outlook / Microsoft 365

1. SMTP Settings

2. Enable SMTP AUTH (Admin)

M365 disables SMTP AUTH by default. An admin must enable it:

1. Go to Microsoft 365 Admin Center → Users → Active Users
2. Select the sending account → Mail → Manage email apps
3. Check Authenticated SMTP and save

Alternative: Microsoft Graph API

For production use, the Graph API with app-only auth is more reliable:

POST https://graph.microsoft.com/v1.0/users/{sender}/sendMail
Authorization: Bearer {token}

{
  "message": {
    "subject": "Weekly Forecast Coaching",
    "body": { "contentType": "HTML", "content": "<h1>Report...</h1>" },
    "toRecipients": [{ "emailAddress": { "address": "leader@company.com" } }]
  }
}

Requires the Mail.Send application permission with admin consent.

Option C: SendGrid (Transactional Email)

Best for production — high deliverability, no per-user mailbox needed, generous free tier (100 emails/day).

1. Get an API Key

1. Sign up at sendgrid.com
2. Verify a Sender Identity (single sender or domain authentication)
3. Go to Settings → API Keys → Create API Key
4. Grant Mail Send permission
5. Copy the key (starts with SG.)

SMTP Settings

Or use the SendGrid API directly

POST https://api.sendgrid.com/v3/mail/send
Authorization: Bearer SG.your-api-key
Content-Type: application/json

{
  "personalizations": [{"to": [{"email": "leader@company.com"}]}],
  "from": {"email": "automation@company.com", "name": "Automation Catalog"},
  "subject": "Weekly Forecast Coaching",
  "content": [{"type": "text/html", "value": "<h1>Report</h1>"}]
}

Platform Configuration

n8n

1. Go to Settings → Credentials → Add Credential
2. Select SMTP (or SendGrid / Gmail for those providers)
3. Enter host, port, username, password from above
4. Use the Send Email node in your workflow

Python (LangGraph / Claude Agent / OpenAI Agent)

export SMTP_HOST="smtp.gmail.com"      # or smtp.office365.com, smtp.sendgrid.net
export SMTP_PORT="587"
export SMTP_USER="you@company.com"      # or "apikey" for SendGrid
export SMTP_PASS="your-app-password"    # or API key for SendGrid

# Python example
import os, smtplib
from email.mime.text import MIMEText
from email.mime.multipart import MIMEMultipart

msg = MIMEMultipart("alternative")
msg["Subject"] = "Weekly Forecast Coaching"
msg["From"] = os.environ["SMTP_USER"]
msg["To"] = "leader@company.com"
msg.attach(MIMEText(plain_text, "plain"))
msg.attach(MIMEText(html_content, "html"))

with smtplib.SMTP(os.environ["SMTP_HOST"], int(os.environ["SMTP_PORT"])) as s:
    s.starttls()
    s.login(os.environ["SMTP_USER"], os.environ["SMTP_PASS"])
    s.send_message(msg)

Make / Power Automate / Zapier

• Make: Email → Send (SMTP), Gmail → Send, or SendGrid → Send
• Power Automate: Send an email (V2) action (native Office 365), or SMTP via HTTP
• Zapier: Email by Zapier (built-in, limited), Gmail integration, or SendGrid integration

Provider Comparison

Troubleshooting