Datasets:
roanbrasil
/
devops-gitops-corpus

Dataset card Data Studio Files
xet
 Community
1
text
stringlengths
0
59.1k
════════════════════════════════════════════
VOLTAGENT SERVER STARTED SUCCESSFULLY
════════════════════════════════════════════
✓ HTTP Server: http://localhost:3141
VoltOps Platform: https://console.voltagent.dev
════════════════════════════════════════════
[VoltAgent] All packages are up to date
```
The [VoltOps Platform](https://console.voltagent.dev) provides debugging and interaction capabilities.
![VoltOps Platform](https://cdn.voltagent.dev/examples/with-ad-generator/voltops-platform.png)
### Agent Architecture
The system consists of three agents:
1. **Supervisor Agent** - Orchestrates the entire workflow and coordinates subagents
2. **Landing Page Analyzer** - Extracts brand information from websites
3. **Ad Creator Agent** - Generates Instagram ads using Google Gemini
VoltAgent agents are autonomous units that execute specific tasks. Each agent has defined responsibilities, tool access, memory persistence, and delegation capabilities. The framework supports multiple LLM providers and includes OpenTelemetry-based observability.
![agent-list](https://cdn.voltagent.dev/examples/with-ad-generator/agents-overall.png)
Agent specifications:
### Landing Page Analyzer Agent
Extracts brand information from websites:
<details>
<summary>Show landing-page-analyzer.agent.ts</summary>
```typescript
import { Agent } from "@voltagent/core";
import { openai } from "@ai-sdk/openai";
import { Memory } from "@voltagent/core";
import { pageNavigateTool } from "../tools/browserbase/page-navigate.tool";
import { pageExtractTool } from "../tools/browserbase/page-extract.tool";
import { pageObserveTool } from "../tools/browserbase/page-observe.tool";
import { screenshotTool } from "../tools/browserbase/screenshot.tool";
export const createLandingPageAnalyzer = (memory: Memory) => {
return new Agent({
name: "LandingPageAnalyzer",
purpose: "Analyze landing pages to extract brand information and marketing insights",
instructions: `You are a landing page analysis expert specializing in extracting brand information for ad creation.
Your primary responsibilities:
1. Navigate to websites and extract comprehensive brand information
2. Identify product names, taglines, and unique value propositions
3. Understand target audience demographics and psychographics
4. Analyze brand voice, tone, and visual style
5. Capture screenshots for visual reference
6. Extract key features, benefits, and differentiators
When analyzing a landing page:
- First navigate to the URL
- Take a screenshot for visual reference
- Extract structured data including:
* Product/Service name
* Main tagline or headline
* Value propositions (primary and secondary)
* Target audience characteristics
* Brand personality and tone
* Key features or benefits
* Call-to-action messages
* Color schemes and visual style notes
- Observe important UI elements like hero sections, CTAs, and feature highlights
Your analysis should be comprehensive enough to inform creative ad generation.
Focus on understanding what makes the brand unique and how it positions itself.
Output format should be structured JSON data that can be easily consumed by other agents.`,
model: openai("gpt-4o-mini"),
tools: [pageNavigateTool, pageExtractTool, pageObserveTool, screenshotTool],
memory,
});
};
```
</details>
![Landing Page Analyzer](https://cdn.voltagent.dev/examples/with-ad-generator/page-analyzer.png)
Functionality:
- Navigates to websites and analyzes content
- Extracts product names, taglines, value propositions
- Identifies target audience and brand attributes
- Captures screenshots for reference
- Outputs structured JSON data
**Tool usage:** BrowserBase integration includes navigation (page loading), screenshot (visual capture), extraction (structured data), and observation (UI element analysis).
### Instagram Ad Creator Agent
Transforms brand data into Instagram ads via Google Gemini: