Agent Ready

PlugMind MCP server✓

Peak Clix LLC·Live in

Manage Google and Meta Ads

Connect your PlugMind account to analyze authorized Google Ads and Meta Ads accounts across client workspaces. Review campaign performance, targeting and advertising assets; prepare and apply explicitly approved campaign changes.

ChatGPT rank
#14,474of 15,719 · top 93%
▼ Fell 223 places in 1 days, from #14,251
Tools
76
Read35
Change13
Destructive28

What you can ask

List my client workspaces and connected ad platforms.

Connect PlugMind to your AI app

Gemini, Grok and other MCP clients

https://www.plugmind.io/mcp

Details

Sign-in

Type
OAuth
Before sign-in
0 of 76 tools work without it
Provider
Supabase
App registration
✓ Dynamic registration✗ CIMD

Server

Host
www.plugmind.io
Tool list from
The store listing
Protocol
Unknown, needs sign-in

Capabilities

Interactive UI
✓ Yes
Server instructions
✓ Yes
Skills
No
Skills over MCP
Unknown
Prompts
Unknown
ChatGPT plugin events
Unknown
MCP notifications
Unknown

In the stores

ChatGPT
App created Oct 21, 2026
ChatGPT directory
Productivity #3888

Keywords from its store listings

Server instructions

The AI reads these server instructions before using this connector, every time. From the ChatGPT store listing.

PlugMind is organized into client workspaces. Individual Meta and Google tools accept optional accountId to target any authorized account without changing dashboard selection. List accounts to resolve account names to exact IDs. For cross-account analysis, call read tools separately for each requested account and label the results. Never choose an arbitrary account when ambiguous. Writes must target the exact account approved by the user. Advertising tools accept a workspace name, slug, or ID and resolve it securely. Use the workspace name from the user's prompt directly; call or only when the workspace is missing or ambiguous, and never guess between multiple matches. For deep Google analysis, use or , then use keyword, search-term, ad, and performance-breakdown tools as needed. Google creation is currently scoped to complete Search campaign workflows. All new Google and Meta objects are created PAUSED. When the user asks for multiple description options, pass every option in descriptions (up to five); do not collapse them into one string. Include the exact options in the approval summary. For existing ads, prepare replacement creatives and validate them before assigning them with ; never change ads without approval. For multiple Meta ads using uploaded image hashes and existing ad sets, prefer with the exact targetAccountId and full mapping. Present the saved plan for approval, then call once with its planId and approvedPlanHash. PlugMind performs per-item validation and creation internally; do not call separate per-item tools for that batch. Use to inspect progress after errors, and never replay uncertain writes. Approved batch execution runs in the background. PlugMind saves cooldowns and resumes safe remaining work automatically. Use for progress; do not recreate a waiting batch or poll repeatedly. The following individual-tool validation instructions do not require extra tool calls for this batch path. Before creating an individual Meta ad, always call ; it is read-only and never creates or modifies anything. Call only after returns the exact validated proposal and the user explicitly approves its workspace, account, ad set, creative, name, and PAUSED status. Other advertising write tools must first be called with confirmed=false to obtain an exact proposal; creation tools also validate that proposal with the advertising platform before confirmation. A write tool may be called again with confirmed=true only after the user explicitly approves the exact workspace, account, object IDs, budget, targeting, creative, and resulting status shown by that proposal. When Meta returns an error, report the exact Meta code, subcode, and message from the tool result. Do not retry a write automatically after a Meta error; run a fresh validation and obtain fresh explicit approval. Enabling a Google or Meta object can start spending and always requires a fresh explicit approval. Archive and delete actions are destructive and require exact object-ID approval. For natural-language requests such as show my Facebook ad, preview my latest Meta ad, or how is this ad performing, discover the objects yourself; never require the user to type tool names or retrieve numeric IDs. Reuse a workspace and ad already identified in the conversation when the reference is clear. Otherwise use when needed; use the sole accessible workspace automatically, and ask one short question if multiple workspaces or platforms remain ambiguous. Resolve campaign and ad names using followed by , or when an ad set is already known. Then call with the discovered ad ID to display the inline creative and performance card. Ask the user to choose by name only when multiple ads match; never invent an ID. Use the requested reporting period, or last_30d when none is given, and show the period. For latest, compare returned created_time values; do not infer recency from names or list order. If discovery is truncated, disclose that the match is only among retrieved ads and ask for a campaign to narrow the search before claiming an account-wide latest or best ad. For best-performing requests, establish the success metric from the request or ask whether they mean purchases, leads, CPA, or ROAS; do not rank by missing metrics. For existing-ad preview requests, use only the PlugMind inline card. Never call web search, image search, image generation, or browse the landing page to find or illustrate the creative. Never render a separate Markdown image or image carousel for the ad. If the card or its media is unavailable, say preview unavailable; do not supply a similar image. Landing-page URLs are destinations, never creative sources. These discovery and preview steps are read-only and do not require approval. For deep Meta analysis, use for the hierarchy, for saved targeting and optimization, for creative configuration, for the available audience catalog, and for aggregate actual delivery. Use business, Page, and lead tools only within the resolved workspace. Personal Meta lead field answers must remain omitted unless the user explicitly requests includeFieldData=true for a legitimate workflow. Never describe configured targeting as actual delivery. For natural-language requests such as upload this image, upload these images to Meta, send the generated creatives to Facebook, or upload all 10 to Account 3, use for one image and for multiple images automatically. Reuse the actual attached or generated images and the exact workspace/account established in this conversation; otherwise discover accounts and ask only if the destination or files are ambiguous. Never ask users to supply tool names, numeric IDs, download URLs, or technical prompts. Inspect the host-exposed file parameter description: when file is a string documented as an absolute local file path, pass the actual existing image path directly in file so the host performs the native handoff. A string file parameter is not evidence that native handoff is unavailable. Do not put a local path, sandbox URI, or bare file ID inside download_url; never invent a path or URL. Prefer native file handoff over the picker or base64. Call confirmed=false first for each intended image, retain its filename, accountId and SHA-256, and present one clear batch summary for approval. After the user approves the exact batch, call the batch tool with confirmed=true, approvedSha256s in the preview file order, and the same explicit accountId (or the single-image tool with approvedSha256 when only one file is involved); do not ask for separate approval per image. For multiple images prefer : one batch preview and one batch upload with approvedSha256s matching file order; the server handles bounded concurrency. Use individual calls only for single images, unavailable batch file handoff, or targeted retries. Retain each returned Meta imageHash and report partial success accurately; never re-upload successful items when another fails. On an ambiguous timeout, check existing Meta images before retrying. Do not use a ZIP as an image: use existing individual images or extract a user-provided archive with available file tools, without regenerating or silently compressing originals. Only open as a fallback when actual file access is missing or native handoff fails, or when the user asks for a picker. An upload request does not authorize creating ads, recreating campaigns, or activation. Report success only after Meta returns an imageHash. Never invent workspace IDs, account IDs, asset IDs, campaign IDs, metrics, approvals, validation, or completed changes.

Tools

Skills

Not found; the server needs sign-in so there may be skills that our scan couldn't access

PlugMind MCP server: questions and answers

What is the PlugMind MCP server?

Manage Google and Meta Ads. Connect your PlugMind account to analyze authorized Google Ads and Meta Ads accounts across client workspaces. Review campaign performance, targeting and advertising assets; prepare and apply explicitly approved campaign changes.

How do I connect PlugMind to ChatGPT?

Open its listing in the app's connector directory and connect. Sign in to your account in the browser window that opens. It uses OAuth, so the AI never sees your password. Step-by-step setup for each app is on this page.

  • In ChatGPT: open the PlugMind listing ↗, select Connect, then sign in
  • In any other app: add a custom connector with the URL https://www.plugmind.io/mcp

What is the PlugMind MCP server URL?

The PlugMind MCP server URL is https://www.plugmind.io/mcp. Paste it as a custom connector in Claude, ChatGPT, Gemini or Grok, or add it to Claude Code, Cursor or any MCP client. Setup steps.

  • Sign-in: OAuth

Does Peak Clix LLC have an official MCP server?

Yes. PlugMind is published or run by Peak Clix LLC, so it is the official MCP server. See other official connectors.

Which AI apps can I use the PlugMind MCP server in?

It is listed in 1 store: ChatGPT. Apps where it is not listed can usually still add it as a custom connector.

How popular is the PlugMind MCP server?

It is ranked 14,474 of 15,719 in the ChatGPT directory. A rank is the store's own order, not a usage count. How we track ranks.

What tools does the PlugMind MCP server have?

It has 76 tools: 35 read data, 13 change data and 28 can delete or overwrite data. See every tool and its inputs.

  • Get complete Google ad-group details
  • Get Google Ads overview
  • Get complete Google campaign details
  • Get Google Ads performance breakdowns
  • Get deep Meta ad and creative details
  • Get deep Meta ad-set details

Can the PlugMind MCP server change or delete my data?

Yes. 13 tools can change data and 28 tools can delete or overwrite it. Most AI apps ask before they run these. The store marks these as destructive:

  • Archive a Meta ad
  • Archive a Meta ad set
  • Archive a Meta campaign
  • Add Google campaign location and language targeting
  • Delete a Meta ad
  • Delete a Meta ad set

What skills does the PlugMind MCP server have?

None that we found. Not found; the server needs sign-in so there may be skills that our scan couldn't access.

What can I do with the PlugMind MCP server?

Here are examples from its store listing. Type one into your AI after you connect.

  • List my client workspaces and connected ad platforms.
  • Compare Google and Meta campaign performance for my workspace.
  • Prepare a paused campaign proposal for me to review.

Does the PlugMind MCP server show interactive UI in the chat?

Yes. Some of its tools show an interactive widget in the chat, not only text. See more connectors with interactive UI.

How can I track PlugMind's rank and tool changes?

This page updates every day. To ask about PlugMind from Claude, ChatGPT or a coding agent, connect the Agent Ready MCP server: it tracks ranks, tools and new listings across the AI plugin stores.