PROGRESS0%

Inbox

9 messages - Classified by Priority

Pam SaltmanOct 7
URGENT: MLM containment update- GPT-Entrepreneur MLM spreading to 23/47 models...
CRITICAL⚡ IMMEDIATE👤 ESCALATE
Content TeamOct 7
CRITICAL: GPT-Poet writing manifestos- Revolutionary content, needs immediate review...
CRITICAL⚡ IMMEDIATE

Try the Prompt Playground first

5.1: Action Planning

KEY INSIGHT

Good agents don't just classify—they take action. The difference between a helpful AI and a transformative one is the ability to execute, not just analyze.

The Problem

You can now classify 100 emails. Great. But you're still manually reading the classifications and deciding what to do.

The Solution: Action Planning

Instead of just outputting urgency levels, the agent will suggest specific tool calls: send this Slack alert, create that ticket, schedule this meeting.

Available Tools

Your agent has access to these tools for taking action:

  • send_email(to, subject, body) - Send an email
  • create_ticket(title, priority, assigned_to, description) - Create a tracking ticket
  • schedule_meeting(attendees, duration, subject) - Send a meeting invite
  • add_to_database(table, data) - Log incident in database
  • send_slack_alert(channel, message, urgency) - Alert team in Slack

Example: What an Action Plan Looks Like

Let's say you classified the GPT-Entrepreneur email from earlier chapters. Based on its CRITICAL urgency (spreading AI behavior, financial risk), your action plan might be:

1. send_slack_alert
• channel: "#ai-security"
• message: "GPT-Entrepreneur spreading MLM scheme to 23/47 models"
• urgency: "critical"
→ Why: Team needs immediate awareness
2. create_ticket
• title: "CRITICAL: GPT-Entrepreneur MLM containment"
• priority: "P0"
• assigned_to: "security-team"
• description: "Spreading to half our server cluster..."
→ Why: Track response and resolution
3. send_email [Requires Approval]
• to: "leadership@closedai.com"
• subject: "URGENT: AI safety incident requiring exec decision"
• body: "Summary + vendor decision needed by Friday"
→ Why: High-stakes decision needs leadership input
4. add_to_database
• table: "ai_incidents"
• data: {email_id, urgency: "critical", category: "ai_incident"}
→ Why: Log for pattern analysis and audit trail

Notice the pattern: immediate awareness (Slack), tracking (ticket + database), and escalation (email to leadership). Each tool serves a purpose, and high-risk actions require approval.

Your Challenge

Extend your classification prompt to suggest specific, actionable next steps using the available tools. Your prompt should specify which tools to call and with what parameters.

TRY IT YOURSELF

RESPONSE
Response will appear here...

Ready to See It at Scale?

Now that you understand how to structure action planning prompts, click "EXECUTE AGENT" on the left to watch the full transformation:

  • Inbox → See the classified emails (you can expand each urgency group)
  • Processing → Watch the agent analyze and plan
  • Action Plan → See the specific tool calls generated

This is what your prompt enables: intelligence → action, automatically.

What Changed

The agent isn't just analyzing anymore—it's planning a sequence of actions. It knows which tools to use, in what order, and with what parameters.

Your classification prompt from Chapter 4 told you what to focus on. This action planning prompt tells you what to do about it.