Build a placements dashboard in Claude

Builds the Get MCP Ads Placements template as a Claude dashboard for Meta, Google Ads, TikTok, Pinterest and LinkedIn: spend, impressions, clicks, conversions and value per placement (Meta publisher platform and position such as Facebook Feed, Instagram Reels or Audience Network; Google Ads network by campaign type, Search partners and Display included, and the sites, apps and channels where Display, YouTube and Demand Gen ads showed; TikTok, Pangle and Global App Bundle; Pinterest Search and Browse; LinkedIn and its Audience Network) with CPM, CTR, CPC, CPA, ROAS and each placement's share of the platform's spend and conversions, against an equivalent comparison period, by day, by campaign and, on Meta, by device and with reach per placement.

Recipe

Intent A live page showing where the ads actually run on each platform and which placements earn their spend, so the team can exclude, shift or keep a placement.

Steps: 3
Tools: 1
Rules: 5
Failure modes: 3

skills/placements-dashboard.md

When to use it

Use it when

Someone asks where their ads run, whether Audience Network, Search partners, the Display network or an app is worth its spend, or how Reels, Feed and Stories compare.
Before excluding a placement or moving budget between placements, and to check afterwards what changed.

Do not use it when

Snapchat placements: Snapchat reports none, say so and answer with its campaigns instead.
Excluding a placement: the dashboard reads only; changes go through the platform tools with preview and confirmation.
Hosts without Claude's Dashboard artifact type: answer in the conversation with the matching reports instead.

What it needs first

Optional Meta Ads, Google Ads, TikTok Ads, Pinterest Ads, LinkedIn Ads

No single source is required: this one reads across whatever you have connected.

Preconditions

The organisation is on Team or Enterprise: ready-made dashboards are not offered on Free or Pro.
The account is among the chosen accounts, confirmed via list_accounts.

Evidence every conclusion must carry

Every figure carries its account, metric and window.
Anything below a readable sample size is reported as unjudged, not as zero.

Inputs
Input	Required	What it is
`window`		Analysis window.Default: last 30 days.

Procedure

In order. Every tool named is one the gateway ships, and links to its reference.

Steps

Resolve the client's workspace and its ad accounts, one per platform (Meta, Google Ads, TikTok, Pinterest, LinkedIn), with list_accounts.
Fetch the placements template with every account (accounts, one per platform, each with its currency).get_dashboard_template
Create a Dashboard artifact, write every document exactly as given in one batch, read its check, hand over with the lines given.

Decision rules

The observable condition, what it lets you conclude, and what takes the conclusion back.

When The template is refused with PLAN_REQUIREDConclude Say the dashboards come with Team and Enterprise and that only the account owner changes the plan; answer the question in the conversation meanwhile.
When The client also advertises on SnapchatConclude Leave Snapchat out of the template and say why: its reporting splits delivery by geography, demographics, device and interests, never by placement.
When The person asks which placement to cutConclude Compare CPA and ROAS within one platform and under one conversion type, weigh the placement's share of spend and conversions, and say that a placement converting little may still feed others (view-through, upper funnel); never compare conversions across platforms.
When A placement shows as UnclassifiedConclude It is delivery the platform reports in its total but not by placement: leave it as is, never spread it over the others.
When A reader sees a source fail to loadConclude Their own Get MCP Ads access decides: check that their connection covers this client's workspace and that their plan includes dashboards before changing the dashboard.

Where the agent stops

The readout changes nothing. Any change it motivates goes through Safe Writes: preview, human confirmation, then apply.

How it goes wrong quietly

The cases where the analysis is wrong and still looks right. Read them before trusting a number.

Editing the template's page or queries by hand loses the provenance of each figure: ask for changes by name instead.
Adding reach across placements or platforms: the same person sees several placements, so reach is only read placement by placement.
Adding conversions across platforms, or treating Google's Performance Max Search traffic as Search campaigns: each row stays within its platform and campaign type.