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

# Agent Architecture

> How Adspirer's AI agent framework uses persistent strategy, multi-account support, and pre-built skills to manage ad campaigns.

# Agent Architecture

Adspirer's agent framework goes beyond raw MCP tools. It provides a structured layer that makes AI assistants into capable performance marketing agents.

## Components

```mermaid theme={"theme":{"light":"github-light","dark":"github-dark"}}
flowchart TD
    A[Agent Instructions] --> B[Strategy Context]
    A --> C[Skills]
    A --> D[MCP Tools]
    B --> E[STRATEGY.md]
    C --> F[Campaign Management]
    C --> G[Performance Review]
    C --> H[Wasted Spend Audit]
    C --> I[Ad Copy Generation]
    D --> J[100+ Platform Tools]
```

### Agent Instructions

Pre-built prompts that configure the AI as a performance marketing expert. Available for Claude Code, Cursor, Codex, and OpenClaw.

### Strategy Context

A persistent `STRATEGY.md` document that captures your brand positioning, target audience, competitive landscape, and campaign history. Every action the agent takes references this context.

### Skills

High-level workflows that combine multiple tools into guided processes. See the [Skills Overview](/skills/overview).

### MCP Tools

The 100+ individual tools for interacting with ad platforms. See platform-specific tool references:

* [Google Ads Tools](/platforms/google-ads)
* [Meta Ads Tools](/platforms/meta-ads)
* [LinkedIn Ads Tools](/platforms/linkedin-ads)
* [TikTok Ads Tools](/platforms/tiktok-ads)