Getting Started Business

Getting Started — For Teams

Welcome to Promptly! This guide walks you through adopting Promptly for AI prompt governance. No technical background required — this is for product managers, team leads, compliance officers, and anyone involved in managing AI behavior.

---

Step 1: Create Your First Project

Everything starts with a Project — your workspace that groups related prompts and controls access.

When to create a project:

• One project per product (e.g., "Customer Support Bot")
• One project per business unit (e.g., "Marketing AI Tools")
• One project per environment (e.g., "Prod — Order Agent")

To create a project:

1. Log in to Promptly.
2. Click the project selector in the top navigation bar.
3. Select "New Project" and give it a name and description.

---

Step 2: Invite Your Team

Promptly uses role-based access to ensure the right people have the right level of control.

Role	Best For
Viewer	Stakeholders who need read-only access.
Editor	Prompt authors — people writing AI instructions.
Reviewer	Team leads or domain experts who approve prompts.
Admin	Project owners who manage settings and membership.

Navigate to Settings → Members to invite team members and assign roles.

---

Step 3: Author Your First Prompt

Navigate to the Prompt Registry and click "New Prompt".

Provide:

• Name — A clear identifier (e.g., refund-policy-agent).
• Content — The instructions for the AI. Use the built-in Monaco Editor.
• Tags — Optional labels for organization (e.g., customer-service, billing).
• Model Configuration — Target AI model and parameters like temperature.

Your prompt is created in Draft status — safe to experiment.

---

Step 4: Scan for Vulnerabilities

Run a Security Scan before submitting for review. The AI-powered scanner checks for:

• 🔓 Prompt injection risks
• 🕵️ PII/PHI data exposure
• 🚧 Missing safety guardrails
• ⚠️ Toxicity and bias concerns

Results:

• Green (Pass): Good to submit.
• Yellow (Warnings): Non-critical recommendations.
• Red (Critical): Must fix before proceeding. Remediation guidance is provided.

---

Step 5: Submit for Review

Click "Submit for Review" to move from DRAFT to IN_REVIEW. Reviewers are notified automatically.

What reviewers see:

• Full prompt content
• A diff viewer showing exactly what changed
• Security scan results

Reviewers can Approve or Reject with comments.

---

Step 6: Deploy

Once approved, deploy the prompt. It becomes immediately available via the Runtime Delivery API.

• No code changes required in your application.
• AI agents automatically use the latest approved version.
• Rollback to any previous version with one click if needed.

---

Step 7: Monitor & Audit

With your prompt live:

• Dashboard — Real-time stats: total prompts, deployed count, prompts in review, security alerts.
• Audit Viewer — Complete history of every action on every prompt.
• Security Scans — Track your security posture over time.

---

The Full Lifecycle

flowchart LR
    A["📝 Create Project"] --> B["👥 Invite Team"]
    B --> C["✏️ Author Prompt"]
    C --> D["🛡️ Security Scan"]
    D --> E["📤 Submit for Review"]
    E --> F["✅ Approve"]
    F --> G["🚀 Deploy"]
    G --> H["📊 Monitor & Audit"]
    H --> C

Loading

---

What's Next?

• Key Concepts — Understand the building blocks of the platform.
• Use Cases — See how teams in different industries use Promptly.
• Why Promptly? — Share the value proposition with stakeholders.
• Platform Overview — Dive deeper into the UI and features.

💡 Tip: You don't need to be a developer to use Promptly. The platform is designed so that product managers, compliance teams, and business stakeholders can participate directly in AI governance.