> ## Documentation Index
> Fetch the complete documentation index at: https://docs.factory.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Droid Shield Plus

> AI-powered security scanning for prompt injection, secrets detection, and sensitive data protection. Enterprise feature powered by Palo Alto Networks Prisma AIRS.

<Info>
  **Enterprise Feature** — Droid Shield Plus is available for enterprise customers. [Contact us](mailto:sales@factory.ai) to enable this feature for your organization.
</Info>

## What is Droid Shield Plus?

Droid Shield Plus is an advanced AI-powered security layer that provides real-time protection against prompt injection attacks, sensitive data exposure, and other security threats. Unlike the standard [Droid Shield](/cli/account/droid-shield) which uses pattern-based detection, Droid Shield Plus leverages **[Palo Alto Networks Prisma AIRS](https://docs.paloaltonetworks.com/ai-runtime-security)** (AI Runtime Security) to provide intelligent, context-aware security scanning.

***

## Key Features

<CardGroup cols={2}>
  <Card title="Prompt Injection Detection" icon="shield-exclamation">
    Identifies and blocks malicious prompt injection attempts designed to manipulate AI behavior or bypass security controls.
  </Card>

  <Card title="Advanced Secrets Scanning" icon="key">
    AI-powered detection of API keys, tokens, passwords, and credentials with higher accuracy and fewer false positives than pattern matching.
  </Card>

  <Card title="Sensitive Data Protection (DLP)" icon="user-shield">
    Detects personally identifiable information (PII), financial data, and other sensitive information before it's exposed in prompts or commits.
  </Card>

  <Card title="Malicious Code Detection" icon="bug">
    Identifies potentially dangerous code patterns and suspicious content that could pose security risks.
  </Card>
</CardGroup>

***

## How Droid Shield Plus Works

Droid Shield Plus provides two layers of protection:

### 1. Prompt Security Scanning

Every prompt you send to Droid is automatically scanned before processing. If a threat is detected, the prompt is blocked and you're notified:

```
Droid Shield Plus has blocked this prompt due to detected: prompt injection, sensitive data.

If you believe this is a false positive, you can disable Droid Shield Plus in settings.
```

**Detected threat categories:**

* **Prompt Injection** — Attempts to manipulate AI instructions
* **Sensitive Data (DLP)** — PII, credentials, or confidential information
* **Toxic Content** — Harmful or inappropriate content
* **Malicious Code** — Suspicious code patterns

### 2. Git Commit Scanning

When you perform `git commit` or `git push` operations through Droid, Droid Shield Plus scans your staged changes using AI-powered analysis:

```
Droid Shield Plus has detected potential secrets in your staged changes.

If you would like to override, you can either:
1. Perform the commit/push yourself manually
2. Disable Droid Shield Plus in settings
```

This provides significantly more accurate detection than regex-based scanning, catching:

* Obfuscated or encoded secrets
* Context-aware credential detection
* Custom secret formats
* Secrets embedded in complex code structures

***

## Droid Shield vs Droid Shield Plus

| Feature                        | Droid Shield           | Droid Shield Plus              |
| ------------------------------ | ---------------------- | ------------------------------ |
| **Detection Method**           | Pattern/Regex matching | AI-powered analysis            |
| **Prompt Scanning**            | No                     | Yes                            |
| **Git Commit Scanning**        | Yes                    | Yes                            |
| **Prompt Injection Detection** | No                     | Yes                            |
| **Sensitive Data (DLP)**       | Limited                | Comprehensive                  |
| **Toxic Content Detection**    | No                     | Yes                            |
| **Malicious Code Detection**   | No                     | Yes                            |
| **False Positive Rate**        | Higher                 | Lower                          |
| **Availability**               | All users              | Enterprise                     |
| **Powered By**                 | Built-in patterns      | Palo Alto Networks Prisma AIRS |

<Note>
  When Droid Shield Plus is enabled, it replaces the standard Droid Shield for git operations, providing enhanced AI-powered scanning instead of pattern-based detection.
</Note>

***

## Enabling Droid Shield Plus

<Steps>
  <Step title="Contact Factory">
    Reach out to [sales@factory.ai](mailto:sales@factory.ai) to enable Droid Shield Plus for your organization.
  </Step>

  <Step title="Enable in Settings">
    Once enabled for your organization:

    1. Run `droid`
    2. Enter `/settings`
    3. Navigate to the **Security** section
    4. Toggle **"Droid Shield Plus (AI-Powered)"** to On
  </Step>

  <Step title="Verify Activation">
    You'll see a subtitle confirming: *"AI-powered security scanning for prompt injection, sensitive data, and toxic content. Powered by Palo Alto Networks Prisma AIRS."*
  </Step>
</Steps>

***

## Handling Blocked Prompts

When Droid Shield Plus blocks a prompt or git operation:

<AccordionGroup>
  <Accordion title="Review the detection">
    Carefully examine what was flagged. The error message will indicate the threat category (prompt injection, sensitive data, toxic content, or malicious code).
  </Accordion>

  <Accordion title="Remove sensitive content">
    If sensitive data was detected:

    * Move secrets to environment variables
    * Use secure credential stores
    * Remove PII from prompts
  </Accordion>

  <Accordion title="Rephrase if needed">
    If prompt injection was detected, rephrase your request to avoid patterns that could be misinterpreted as manipulation attempts.
  </Accordion>

  <Accordion title="Report false positives">
    If you believe the detection is incorrect, contact [support@factory.ai](mailto:support@factory.ai) with details about the false positive.
  </Accordion>
</AccordionGroup>

<Warning>
  **Never disable Droid Shield Plus just to bypass security checks.** If content is being blocked, there's likely a legitimate security concern that should be addressed.
</Warning>

***

## Security & Privacy

Droid Shield Plus is powered by **[Palo Alto Networks Prisma AIRS](https://docs.paloaltonetworks.com/ai-runtime-security)**, a leading AI runtime security platform:

* **Real-time Analysis** — Prompts and code are scanned in real-time with minimal latency
* **Enterprise-Grade Security** — Built on Palo Alto Networks' industry-leading security infrastructure
* **Graceful Degradation** — If the security service is temporarily unavailable, Droid continues to function while logging the issue

***

## Related Resources

<CardGroup cols={2}>
  <Card title="Droid Shield" icon="shield" href="/cli/account/droid-shield">
    Learn about the standard pattern-based secret detection included with all Droid accounts.
  </Card>

  <Card title="Security Overview" icon="lock" href="/cli/account/security">
    Comprehensive overview of Factory's security features and practices.
  </Card>
</CardGroup>

### Prisma AIRS Documentation

<CardGroup cols={2}>
  <Card title="Prisma AIRS Overview" icon="book" href="https://docs.paloaltonetworks.com/ai-runtime-security">
    Official Palo Alto Networks documentation for Prisma AIRS AI Runtime Security platform.
  </Card>

  <Card title="Prisma AIRS API Reference" icon="code" href="https://pan.dev/prisma-airs/api/airuntimesecurity/">
    Developer documentation for the Prisma AIRS AI Runtime Security API.
  </Card>
</CardGroup>

***

## Get Droid Shield Plus

<Card title="Enable for Your Organization" icon="building" href="mailto:sales@factory.ai">
  Contact our sales team at **[sales@factory.ai](mailto:sales@factory.ai)** to enable Droid Shield Plus for your enterprise organization.
</Card>