AdsAgent Skill Pack (docs)
OfficialRelated Servers
Alternatives to AdsAgent Skill Pack (docs)
No user-submitted related servers found.
Related Servers
- AlicenseNot gradedqualityBmaintenanceEnables AI assistants to query and manage Google Ads accounts, covering reporting metrics like ROAS, CTR and CPC, custom GAQL queries, keyword research, and campaign management such as creating campaigns, ad groups, ads and targeting. It connects Claude, ChatGPT, Cursor and other MCP clients to Google Ads data through a hosted remote server with OAuth.MIT
- AlicenseNot gradedqualityAmaintenanceEnables AI agents to create, manage, and automate Meta (Facebook/Instagram) ads campaigns via the Marketing API, with safe defaults and AI-generated creative.1MIT
- AlicenseNot gradedqualityCmaintenanceEnables AI agents to run Meta (Facebook and Instagram) advertising: pull spend, CPA, ROAS and breakdown reports, audit account health, build and boost campaigns, ad sets and ads, and pause or adjust budgets, all starting paused until the user approves going live. Also supports brand memory, change logs and an optional daily spend cap.MIT
- AlicenseAqualityAmaintenanceMCP server for AI agents to manage ad campaigns across Google, Meta, LinkedIn, Microsoft, Reddit, TikTok, and more21175 npm21MIT
- AlicenseCqualityAmaintenanceEnables any MCP-compatible AI client to manage Reddit Ads accounts through the Reddit Ads API v3, creating, updating, deleting and reporting on campaigns, ad groups, ads, creatives, targeting, audiences, catalogs and conversions. It handles browser-based OAuth login, pagination, token refresh and rate-limit backoff automatically.11424 npm2MIT
- AlicenseNot gradedqualityCmaintenanceEnables AI clients to read Meta Ads performance with breakdowns and diagnostics and to create, update, or pause campaigns, ad sets, ads, creatives, audiences, and catalogs, with every write requiring explicit user approval. Runs as a hosted, OAuth-authenticated remote server with no local install.1MIT
TDQS
Scored across 4 tools
Each tool has a clear, distinct responsibility: list_skills catalogs available skills, get_skill retrieves a specific skill's documentation, get_pack_readme returns pack identity/version, and get_hosted_mcp_urls exposes hosted endpoint URLs. There is no meaningful overlap or ambiguity between them.
Tool names follow a predictable verb_noun pattern: get_* for single-item retrieval and list_skills for collection retrieval. The naming is consistent, descriptive, and easy to infer.
Four tools is well-scoped for a documentation/skill-pack server. Each tool earns its place and covers a necessary retrieval function without unnecessary bloat.
The tool surface covers the full documentation workflow: discover available skills, fetch individual skill details, confirm pack identity/version, and retrieve hosted MCP URLs. There are no obvious dead ends for the stated documentation-only purpose.