> ## Documentation Index > Fetch the complete documentation index at: https://docs.adsmcp.si/llms.txt > Use this file to discover all available pages before exploring further. # ADSMCP Skill Reference: 13 AI Ad Skills > Reference for all 13 ADSMCP skills — the agent persona, the tool-call contract, six platform skills, four cross-platform workflows, and docs access. ADSMCP ships **13 skills** — instruction files that teach an AI assistant how to run advertising campaigns properly. They follow the open [Agent Skills](https://agentskills.io) format, so the same files work in Claude, ChatGPT, Codex, Cursor, Gemini CLI, and OpenClaw. You don't invoke them by name. Describe what you want — "how are my Google Ads doing?", "launch a TikTok campaign" — and your assistant loads the right skill. > **[Get your free paid media skill](https://docs.adsmcp.si/agent-skills/free-paid-media-skill.md)** > > Download one Markdown file with agent instructions and eight starter prompts for Claude, ChatGPT, or your coding agent. ## How the skills fit together Skills load on demand. Your assistant sees only each skill's name and description until a task matches, then reads the full instructions. Two skills underpin everything: * **`adsmcp-agent`** sets the safety rules — campaigns are created paused, nothing spends without your explicit approval — and routes to the right skill. * **`adsmcp-mcp`** carries the tool-call contract: how to reach each platform's tools, which account id each platform expects, and the budget units (Google's in dollars, Meta's in the ad account's own currency — not cents). Everything else builds on those two. A workflow skill like "launch a campaign" delegates to the platform skill for that platform's field rules, rather than duplicating them. ## Quick reference ### Foundational | Skill | What it does | | - | - | | **`adsmcp-agent`** | Paid-media agent behavior. The safety contract and the routing table. | | **`adsmcp-mcp`** | How to call the tools: the router two-step, account ids, budget units, quota. | ### Per platform | Skill | What it covers | | - | - | | **`adsmcp-google-ads`** | Search, Performance Max, Demand Gen, YouTube, Display. Keywords, extensions, bidding. | | **`adsmcp-meta-ads`** | Facebook and Instagram. Image, video, carousel, lead-gen. Audiences and creative fatigue. | | **`adsmcp-tiktok-ads`** | In-feed video, Spark Ads, carousel, app promotion. The pixel and objective rules. | | **`adsmcp-linkedin-ads`** | Sponsored content and lead-gen, with B2B job-title and company targeting. | | **`adsmcp-amazon-ads`** | Sponsored Products, Brands, and Display. ASIN targeting and ACoS. | | **`adsmcp-chatgpt-ads`** | Ads that appear inside ChatGPT responses. | ### Cross-platform workflows | Skill | What it does | | - | - | | **`adsmcp-launch`** | Plan and launch a campaign. Picks the platform for the goal, splits budget, verifies the build. | | **`adsmcp-performance-review`** | Cross-platform scorecard, plus a conversion-tracking audit. Read-only. | | **`adsmcp-optimize`** | Find wasted spend, add negatives, exclude placements, reallocate budget, check pacing. | | **`adsmcp-creative`** | Write headlines and ad copy. Diagnose and fix creative fatigue. | ### Product questions | Skill | What it does | | - | - | | **`adsmcp-docs`** | Answers questions about ADSMCP — plans, quota, connecting accounts — from these docs. | ### Workspace (local assistants only) | Skill | What it does | | - | - | | **`adsmcp-setup`** | Bootstraps a brand workspace: scans your folder, pulls live campaign data, writes context files. | `adsmcp-setup` needs a filesystem, so it ships to Claude Code, Cursor, Codex, and Gemini CLI — not to the ChatGPT plugin. *** ## What the skills guarantee Every skill inherits the same safety contract from `adsmcp-agent`: > **Nothing spends without your approval** > > Campaigns are created **paused** on every platform. The agent never resumes a campaign, raises a > budget, or changes a bid strategy without you agreeing to that specific action first. * **Read before write.** The agent looks at what's actually in the account before proposing a change. * **Verify after write.** After creating or updating anything, it reads the object back and reports what exists — not what it asked for. * **Never invent a number.** If a tool fails, the agent says so. It does not estimate metrics. * **Destructive means destructive.** Deleting a creative or removing keywords is named as such, and confirmed, before it happens. *** ## Using your assistant's own features The skills know what the app they're running in can do, and offer it when it helps. ### Showing work In **Claude**, a scorecard, a set of ad-copy variants, or a campaign plan can be published as an **artifact** — a live page that updates as work continues, private to you by default. In **ChatGPT**, something you'll revisit or send to a client can be a hosted **Site**. The agent defaults to invitation-only, because these pages carry your spend and conversion data. ### Repeating work Ask for a weekly review or a pacing check and the agent offers to schedule it, using whatever your assistant calls it: | Assistant | What it's called | Runs when your computer is off? | | - | - | - | | ChatGPT | Scheduled tasks | Yes | | Codex | Scheduled tasks (automations) | Yes | | Claude Cowork | Scheduled tasks, via `/schedule` | Yes — they run remotely | | Claude Code | Routines — a *local* task or a *remote* routine | Local: no. Remote: yes | > [!WARNING] > > A Claude Code **local** scheduled task only runs while the desktop app is open and your computer is > awake. A laptop asleep at 9am skips the run. For budget pacing or spend alerts, the agent sets up a > **remote routine** instead. Every scheduled run calls ADSMCP tools again, so it draws on your monthly tool-call quota exactly like a live conversation. The agent tells you what a cadence will cost before setting it up — weekly is the sensible default for a performance review. If your assistant has no scheduler, ADSMCP's own monitoring tools cover the same ground server-side: threshold alerts on ROAS, CPA and spend, and scheduled reports delivered by email. *** ## Deeper reference material Some skills carry `references/` files that load only when the agent needs them — keeping the day-to-day context small: | Reference | Lives in | Why | | - | - | - | | Objective × pixel validity matrix | `adsmcp-tiktok-ads` | TikTok rejects invalid combinations rather than correcting them. | | Creative specs and ODAX objectives | `adsmcp-meta-ads` | Image ratios, carousel rules, character limits. | | Campaign types, bidding, PMax assets, extensions | `adsmcp-google-ads` | Choosing the right campaign and bid strategy. | | Error catalog | `adsmcp-mcp` | Known failures and their fixes, so the agent doesn't retry a call that cannot succeed. | | Docs map | `adsmcp-docs` | Topic-to-URL index of this documentation. | *** ## Getting the skills > **[Claude Code & Cowork](https://docs.adsmcp.si/ai-clients/claude-code.md)** > > Install the ADSMCP plugin from the marketplace. > **[ChatGPT](https://docs.adsmcp.si/ai-clients/chatgpt.md)** > > Skills ship with the ADSMCP app. > **[Codex](https://docs.adsmcp.si/ai-clients/codex.md)** > > Install the ADSMCP plugin. > **[Cursor](https://docs.adsmcp.si/ai-clients/cursor.md)** > > Install the ADSMCP plugin from the Cursor Marketplace. > **[Grok Bot](https://docs.adsmcp.si/ai-clients/grok.md)** > > Add the ADSMCP plugin from the sidebar. Source lives in [github.com/amekala/ads-mcp](https://github.com/amekala/ads-mcp) under `shared/skills/`, and fans out to each client from there. ## Going further: a skill for your own account These 13 skills know advertising. They don't know *your* account — your campaigns, your spend, your currency, what state the account is in. The [**Prompt Engineering Playbook**](https://docs.adsmcp.si/agent-skills/prompt-engineering-playbook.md) is the paid layer that does: a **one-time \$39 per ad account**, and ADSMCP generates a custom skill, an account dossier, a prompt book, and paste-ready agent instructions from that account's real data. You download the files and install them alongside these free skills. | | These 13 skills | Prompt Engineering Playbook | | - | - | - | | **Scope** | Advertising in general | One specific ad account | | **Knows your campaigns** | No | Yes | | **Cost** | Free on every plan | \$39 once, per account (one free on annual) | You want both — the free skills are the grammar, the playbook is the briefing. Launch a Google Search campaign end to end. 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. See the [Tool Catalog](https://docs.adsmcp.si/agent-skills/tools.md) for the tools these skills drive, and [Core Workflows](https://docs.adsmcp.si/agent-skills/workflows.md) for the step-by-step patterns. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.