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

# Adspirer Skill Reference: 5 AI Ad Workflows

> Reference for all 5 Adspirer skills — proven workflows that teach AI how to manage ad campaigns, research keywords, write ad copy, and optimize spend.

Adspirer ships 5 skills — instruction files that teach AI assistants proven advertising workflows. You can invoke them directly or just describe what you want in natural language.

## Quick Reference

| Skill                      | Claude Code                        | Cursor                         | Codex                          | What It Does                                            |
| -------------------------- | ---------------------------------- | ------------------------------ | ------------------------------ | ------------------------------------------------------- |
| **Ad Campaign Management** | `/adspirer:ad-campaign-management` | `/adspirer-ads`                | `$adspirer-ads`                | Full campaign management — all platforms, all workflows |
| **Setup**                  | `/adspirer:setup`                  | `/adspirer-setup`              | `$adspirer-setup`              | Bootstrap a brand workspace                             |
| **Performance Review**     | `/adspirer:performance-review`     | `/adspirer-performance-review` | `$adspirer-performance-review` | Cross-platform performance scorecard                    |
| **Write Ad Copy**          | `/adspirer:write-ad-copy`          | `/adspirer-write-ad-copy`      | `$adspirer-write-ad-copy`      | Brand-voice ad copy from real data                      |
| **Wasted Spend**           | `/adspirer:wasted-spend`           | `/adspirer-wasted-spend`       | `$adspirer-wasted-spend`       | Find and fix wasted ad spend                            |

<Tip>
  You don't need to memorize these commands. Just describe what you want — "how are my Google Ads doing?" or "create a LinkedIn campaign" — and your AI assistant will match the right skill automatically.
</Tip>

***

## Skill Details

### Ad Campaign Management

The main skill. Covers campaign creation, performance analysis, keyword research, budget optimization, ad extensions, audience analysis, monitoring, and reporting across all 6 platforms.

**When to use:** Any advertising task — creating campaigns, analyzing performance, researching keywords, optimizing budgets, writing ad copy, managing ad extensions, or switching accounts.

**What happens when you invoke it:**

1. Checks which ad platforms are connected (`get_connections_status`)
2. Identifies the task type (performance analysis, campaign creation, keyword research, etc.)
3. Follows the platform-specific workflow (see [Core Workflows](/agent-skills/workflows))
4. Loads strategy directives from `STRATEGY.md` if they exist
5. Presents results with recommendations
6. Asks for confirmation before any action that affects spend

<Prompt description="Full campaign creation with keyword research and ad extensions." actions={["copy", "cursor"]}>
  Create a Google Search campaign for my SaaS product:

  * Product: cloud inventory management for restaurants
  * Target: restaurant owners in Chicago
  * Budget: \$75/day
  * Research keywords first, then create the campaign with sitelinks and callout extensions
</Prompt>

This skill contains the complete [Tool Catalog](/agent-skills/tools) (340+ tools) and all [Core Workflows](/agent-skills/workflows).

***

### Setup

Bootstraps a brand workspace from scratch. Scans your folder, connects to your ad accounts, pulls live campaign data, and creates context files that power all other skills.

**When to use:** First time in a new brand folder, or when you want to refresh your brand context after connecting new platforms or adding new docs.

**What happens when you invoke it:**

1. Connects to Adspirer MCP (triggers OAuth if needed)
2. Scans local files for brand docs (`.md`, `.txt`, `.csv`, `.yaml`, `.json`, `.pdf`)
3. Pulls live data: connections, campaigns, performance, search terms, benchmarks
4. Creates the brand context file with:
   * Brand overview, voice, target audiences
   * Connected platforms and active campaign counts
   * Budget guardrails and KPI targets
   * Performance snapshot and key findings
5. Creates `STRATEGY.md` for persisting strategic decisions
6. Presents a summary of what it found

<Prompt description="Bootstrap a brand workspace from scratch." actions={["copy", "cursor"]}>
  Set up my brand workspace
</Prompt>

**What it creates:**

| File                                   | Purpose                                             |
| -------------------------------------- | --------------------------------------------------- |
| `CLAUDE.md` / `BRAND.md` / `AGENTS.md` | Brand context (varies by IDE)                       |
| `STRATEGY.md`                          | Strategy directives that persist across sessions    |
| `MEMORY.md`                            | Past decisions and learnings (Claude Code + Cursor) |

<Tip>
  Drop brand docs (guidelines, media plans, briefs) into your folder before running setup. The more context you provide, the better the agent's ad copy and recommendations.
</Tip>

***

### Performance Review

Runs a cross-platform performance scorecard. Pulls metrics from all connected platforms, compares against KPI targets, and identifies top performers, underperformers, and wasted spend.

**When to use:** Weekly check-ins, monthly reviews, or whenever you want a snapshot of how campaigns are doing across all platforms.

**What happens when you invoke it:**

1. Reads your brand context file for KPI targets
2. Reads `STRATEGY.md` for active directives — flags "Strategy Drift" if campaigns violate them
3. Pulls last 30 days of performance from all connected platforms
4. Pulls wasted spend analysis per platform
5. Presents a unified scorecard with actuals vs. targets
6. Highlights top performers, underperformers, and strategy alignment
7. Recommends top 3 actions

<Prompt description="Cross-platform performance review with strategy alignment." actions={["copy", "cursor"]}>
  Run a performance review for the last 30 days across all platforms. Compare against my KPI targets and flag any campaigns that conflict with my strategy.
</Prompt>

**Output format:**

| Platform | Campaigns | Spend | CTR | CPA | ROAS | Waste | Health |
| -------- | --------- | ----- | --- | --- | ---- | ----- | ------ |
| Google   | ...       | ...   | ... | ... | ...  | ...   | ...    |
| LinkedIn | ...       | ...   | ... | ... | ...  | ...   | ...    |
| Meta     | ...       | ...   | ... | ... | ...  | ...   | ...    |

***

### Write Ad Copy

Generates brand-voice ad copy from real keyword and performance data. Reads your brand guidelines, checks current campaigns, analyzes what users search for, and produces copy that matches your voice and targets high-performing keywords.

**When to use:** When you need new headlines, descriptions, or ad creative text for any platform.

**What happens when you invoke it:**

1. Reads brand context for voice rules and target audiences
2. Reads `STRATEGY.md` for competitive positioning and creative direction
3. Pulls current ad copy and keywords from your campaigns
4. Analyzes search terms to understand user intent
5. Generates AI-powered suggestions from real data
6. Filters through brand voice rules
7. Presents 5+ headline/description options with:
   * Character count validation
   * Keyword theme coverage
   * Pass/fail vs. platform limits
8. Gets your approval before applying changes

<Prompt description="Brand-voice ad copy with keyword coverage validation." actions={["copy", "cursor"]}>
  Write new Google Ads headlines and descriptions for my top-performing campaign. Use my brand voice from BRAND.md and make sure headlines cover the top keyword themes.
</Prompt>

***

### Wasted Spend

Audits all connected platforms for wasted ad spend. Finds zero-conversion keywords, fatigued creatives, underperforming audiences, and budget misallocations. Automatically detects campaigns in learning phase or with insufficient data and excludes them from waste totals.

**When to use:** Monthly budget reviews, when CPA is climbing, or when you suspect money is being wasted on underperforming campaigns.

**What happens when you invoke it:**

1. Checks connected platforms
2. Reads `STRATEGY.md` — flags campaigns violating strategy directives as priority waste
3. Runs platform-specific waste analysis:
   * Google: `analyze_wasted_spend` + `analyze_search_terms`
   * Meta: `analyze_meta_wasted_spend`
   * LinkedIn: `analyze_linkedin_wasted_spend`
4. Separates campaigns into three categories:
   * **Learning phase** (\< 14 days of data) — shown separately, excluded from waste totals
   * **Insufficient data** (\< \$20 spend) — shown separately, excluded from waste totals
   * **Established campaigns** — full ROAS-based waste analysis
5. Runs budget optimization analysis
6. Presents total wasted spend, top waste sources, and potential monthly savings
7. Recommends fixes: negative keywords, audience exclusions, budget shifts
8. Gets your approval before making changes

<Note>
  New campaigns are never flagged as "wasted spend." The tool waits for 14 days of data and at least \$20 in spend before judging performance, so you won't get false alarms on campaigns that just launched.
</Note>

<Prompt description="Cross-platform wasted spend audit with savings estimate." actions={["copy", "cursor"]}>
  Find wasted spend across all my platforms. Show me which keywords have spend but zero conversions, which Meta ads are fatigued, and how much I could save per month.
</Prompt>

***

## Customizing Skills

Skills are plain markdown files (`SKILL.md`). You can edit them to:

* Add custom workflows for your specific use cases
* Modify safety rules (e.g., change minimum budget thresholds)
* Add industry-specific guidance (e.g., healthcare compliance, financial disclaimers)
* Remove sections you don't need

Changes take effect immediately (or after restarting your AI client).

**Skill file locations:**

| IDE         | Location                                           |
| ----------- | -------------------------------------------------- |
| Claude Code | `~/.claude/skills/ad-campaign-management/SKILL.md` |
| Cursor      | `~/.cursor/skills/adspirer-*/SKILL.md`             |
| Codex       | `~/.agents/skills/adspirer-*/SKILL.md`             |
| OpenClaw    | Bundled with plugin (not editable)                 |

## FAQ

<AccordionGroup>
  <Accordion title="Do I need all 5 skills?">
    No. The **Ad Campaign Management** skill is the main one — it covers everything. The other 4 are shortcuts for common tasks. If you only install one skill, install that one.
  </Accordion>

  <Accordion title="What's the difference between a skill and a slash command?">
    They're the same thing. A **skill** is the instruction file (`SKILL.md`) that teaches the AI a workflow. A **slash command** (like `/adspirer:setup`) is how you invoke that skill. In Cursor, skills are invoked with `/adspirer-*`. In Codex, they use `$adspirer-*`.
  </Accordion>

  <Accordion title="Can I create my own skills?">
    Yes. Create a `SKILL.md` file with frontmatter (`name`, `description`) and markdown instructions. Place it in your IDE's skills directory. See the [Adspirer skills source code](https://github.com/amekala/ads-mcp/tree/main/shared/skills) for examples of the format.
  </Accordion>

  <Accordion title="Do skills work across all IDEs?">
    The same skill content works in Claude Code, Cursor, and Codex — but with IDE-specific adaptations. Claude Code uses `CLAUDE.md` for brand context, Cursor uses `BRAND.md`, and Codex uses `AGENTS.md`. The [sync script](https://github.com/amekala/ads-mcp/blob/main/scripts/sync-skills.sh) generates IDE-specific versions from shared templates.
  </Accordion>

  <Accordion title="Skills can't reach Adspirer — what do I do?">
    If skills fail because the MCP server is unreachable, the issue is your connector, not the skill. Follow these steps:

    1. **Check tool permissions** in your AI client's connector settings — read tools should be **Always allow**, write tools should be **Custom**
    2. **Disconnect and reconnect** the Adspirer connector, then complete the OAuth flow again
    3. **Refresh your Adspirer session** — log out of [adspirer.ai](https://adspirer.ai) and log back in

    Claude and ChatGPT web connectors may disconnect every 1–2 weeks. This is normal — just re-enable and re-authenticate. See [Troubleshooting FAQ](/knowledge-base/faq#troubleshooting) for full details.
  </Accordion>
</AccordionGroup>

## Related Documentation

* [Skills Overview](/agent-skills/overview) — What skills are and how to install them
* [Performance Marketing Agent](/agent-skills/agent) — The agent that orchestrates skills with brand knowledge and strategy
* [Core Workflows](/agent-skills/workflows) — Step-by-step tool sequences
* [Tool Catalog](/agent-skills/tools) — All 340+ Adspirer tools