Detect AI Agent Environment

Detect if your code is being executed by an AI agent or automated development environment using the `@vercel/detect-agent` package. This enables adaptive behavior, improved telemetry, and conditional features based on execution context.

What This Skill Does

This skill helps you integrate AI agent detection into Node.js applications, allowing your code to:

Instructions

Follow these steps to implement AI agent detection in the user's project:

1. Install the Package

Install `@vercel/detect-agent` in the project:

```bash

npm install @vercel/detect-agent

```

2. Implement Basic Detection

Create or update the relevant file where agent detection is needed. Import and use the `determineAgent` function:

```typescript

import { determineAgent } from '@vercel/detect-agent';

const { isAgent, agent } = await determineAgent();

if (isAgent) {

console.log(`🤖 Running in ${agent.name} environment`);

// Adapt behavior for AI agent context

} else {

console.log('Running in standard environment');

}

```

3. Apply Context-Appropriate Use Cases

Based on the user's requirements, implement one or more of these patterns:

#### Adaptive Logging Behavior

```typescript

import { determineAgent } from '@vercel/detect-agent';

async function setupEnvironment() {

const { isAgent, agent } = await determineAgent();

if (isAgent) {

// Increase verbosity for AI agents

process.env.LOG_LEVEL = 'verbose';

process.env.OUTPUT_FORMAT = 'structured';

console.log(`🤖 Detected AI agent: ${agent.name}`);

} else {

process.env.LOG_LEVEL = 'info';

}

}

// Call during application initialization

await setupEnvironment();

```

#### Telemetry and Analytics

```typescript

import { determineAgent } from '@vercel/detect-agent';

async function trackUsage(event: string, data?: Record<string, any>) {

const result = await determineAgent();

await analytics.track(event, {

...data,

agent: result.isAgent ? result.agent.name : 'human',

isAutomated: result.isAgent,

timestamp: Date.now(),

});

}

// Usage

await trackUsage('feature_used', { feature: 'deployment' });

```

#### Feature Flags for AI Agents

```typescript

import { determineAgent } from '@vercel/detect-agent';

async function shouldEnableFeature(feature: string): Promise<boolean> {

const result = await determineAgent();

// Enable specific features for AI agents

const aiOnlyFeatures = [

'experimental-ai-mode',

'verbose-error-traces',

'auto-suggestions',

];

if (result.isAgent && aiOnlyFeatures.includes(feature)) {

return true;

}

// Regular feature flag logic

return checkFeatureFlag(feature);

}

// Usage

if (await shouldEnableFeature('experimental-ai-mode')) {

enableExperimentalMode();

}

```

#### Conditional CLI Output

```typescript

import { determineAgent } from '@vercel/detect-agent';

async function displayOutput(message: string, data?: any) {

const { isAgent } = await determineAgent();

if (isAgent) {

// Structured output for AI parsing

console.log(JSON.stringify({ message, data, timestamp: new Date().toISOString() }));

} else {

// Human-friendly output

console.log(message);

if (data) console.log(data);

}

}

```

4. Support the AI_AGENT Standard (Optional)

If the user is building an AI tool, advise them to set the `AI_AGENT` environment variable:

```bash

In their tool's runtime

export AI_AGENT="tool-name"

Or when executing commands

AI_AGENT="[email protected]" node script.js

```

**Naming convention:**

5. Test the Implementation

Create a test script to verify detection works:

```typescript

import { determineAgent } from '@vercel/detect-agent';

async function testDetection() {

const result = await determineAgent();

console.log('Detection Result:', {

isAgent: result.isAgent,

agentName: result.isAgent ? result.agent.name : 'none',

});

}

testDetection();

```

Run the test:

Supported AI Agents

The package detects these environments:

Important Notes

Constraints

Example Use Cases

1. **CLI Tools**: Adjust output verbosity and format when run by AI agents

2. **Build Systems**: Enable additional logging for debugging AI-driven builds

3. **Development Servers**: Provide structured responses for AI consumption

4. **Testing Frameworks**: Tag test runs with agent context for analytics

5. **Deployment Scripts**: Adapt confirmation prompts when automated

Links