Windsor.ai
Windsor.ai connects marketing and business data sources for analytics, reporting, account management, and supported write actions.
Verdict
Common use cases
- Compare ad spend across Google and Meta daily
- Pull yesterday's conversion data for standup
- List which client accounts are connected
- Blend Shopify revenue with Facebook ad cost
- Discover queryable fields before building reports
Integration
- Vendor
- Windsor.ai
- Category
- developer-tools
- Auth
- API_KEY
- Tools
- 6
- Composio slug
windsor_ai
Tools
- List Connector Actions
Discover Windsor.ai write-action IDs and parameter schemas without executing them. Use 'all' to inspect the currently bounded cross-connector catalog.
- List Connector Fields
List queryable field IDs for one specific Windsor.ai connector. This tool rejects connector='all' because that response is unbounded and very large.
- List Connector Options
List connector-specific query options for one connected Windsor.ai source. Returns a clear error when the source is not connected.
- List Connectors
List Windsor.ai connector IDs that can be used with discovery, data-query, and account-connection tools.
- List Co-User Linked Accounts
List source accounts connected through Windsor.ai co-user authorization links, optionally filtered to one connector, without exposing link access tokens.
- Query Connector Data
Query a selected Windsor.ai connector, or 'all' for blended data, using explicit fields and bounded JSON output. Use connector discovery tools first when field or option IDs are unknown.
Setup
Setup guide
- 11. In Switchy, open your workspace settings and navigate to the MCP integrations page. 2. Click 'Add Integration' and select Windsor.ai from the developer-tools category. 3. You'll be prompted to paste your Windsor.ai API key, which you generate in the Windsor.ai dashboard under Settings > API Keys. 4. Grant read access to connector data and linked accounts when prompted. 5. Click 'Connect' and wait for the green confirmation badge. 6. Open any Project, type '@Windsor.ai list connectors' in the chat, and press send — you should see a list of available data sources (google_ads, facebook, shopify, etc.). 7. If the list is empty, connect at least one advertising or analytics platform in your Windsor.ai dashboard, then retry the command. 8. To test a real query, try '@Windsor.ai query connector data from google_ads with fields spend, clicks for the last 7 days' and confirm you receive JSON output.
What teammates see: by default, memories from Windsor.ai are scoped to the Project (PROJECT visibility) - you can mark any memory PRIVATE or share it ORG-wide.
Works well with
Top models
Compatibility data appears once enough Projects have used this MCP together with a given model.
How Switchy teams use it
Starter prompts
Weekly Spend Rollup
@Windsor.ai query all connectors for the field 'spend' over the last 7 days and return the totals by source.
List Connected Accounts
@Windsor.ai list all co-user linked accounts and show the connector name and account ID for each.
Google Ads Performance
@Windsor.ai query the google_ads connector for fields 'clicks', 'impressions', 'cost' for yesterday.
Discover Facebook Fields
@Windsor.ai list all queryable fields for the facebook connector and show their IDs.
Shopify Revenue This Month
@Windsor.ai query the shopify connector for the field 'revenue' from the first of this month to today.
Example outputs
Illustrative - representative of the model's voice and quality, not literal recordings.
@windsor.ai query all connected ad platforms for spend, impressions, and conversions over the last 7 days, grouped by campaign
Windsor.ai returned blended metrics across 4 connected platforms (Google Ads, Meta Ads, LinkedIn Ads, TikTok Ads). Total spend: $12,847; impressions: 2.1M; conversions: 1,203. Top campaign by ROAS: 'Q1_Retargeting' (Meta) at 4.2x. The response includes per-platform breakdowns and daily trend arrays. Note: Currency is normalized to USD; conversion definitions vary by platform and may require manual reconciliation for attribution modeling.
This example showcases Windsor.ai's core strength: unified cross-platform reporting without switching dashboards. Requires API_KEY auth and at least one connected ad account per platform. Be aware that field availability depends on which connectors are linked—use List Connector Fields first if you're unsure what metrics a specific platform exposes.
@windsor.ai list all available connectors and show which ones are currently connected to my account
Windsor.ai supports 45 connectors spanning ad platforms, analytics tools, and CRM systems. Currently connected: Google Ads (3 accounts), Meta Ads (2 accounts), Shopify (1 account), HubSpot (1 account). Not connected but available: LinkedIn Ads, TikTok Ads, Salesforce, Google Analytics 4, Stripe, and 36 others. Each connector exposes different field sets—use List Connector Fields to see queryable dimensions and metrics before building reports.
This read-heavy example demonstrates Windsor.ai's discovery layer, which helps users understand their integration surface before querying. Useful when onboarding or auditing data access. The connector list is static, but connection status reflects real-time account links. No side effects—purely informational.
@windsor.ai compare last month's Google Ads performance to Meta Ads performance, then draft a one-paragraph executive summary highlighting which platform delivered better cost-per-acquisition
Windsor.ai returned: Google Ads spent $8,200 with 340 conversions (CPA: $24.12); Meta Ads spent $6,950 with 298 conversions (CPA: $23.32). Meta Ads delivered a 3.3% lower CPA despite 12% fewer conversions, suggesting better audience targeting efficiency. Draft summary: 'In the prior month, Meta Ads outperformed Google Ads on cost-efficiency with a $23.32 CPA versus $24.12, though Google Ads drove 14% more total conversions. Consider reallocating budget to Meta for bottom-funnel campaigns while maintaining Google's volume advantage for top-of-funnel reach.'
This synthesis example pairs Windsor.ai's cross-platform data retrieval with AI reasoning to produce actionable insights. The MCP fetches raw metrics; the AI interprets trends and drafts narrative summaries. Requires connected accounts for both platforms and assumes conversion tracking is configured consistently. Rate limits apply to bulk historical queries—stay within Windsor.ai's API tier allowances.
Use-case deep-dives
When Windsor.ai wins for multi-channel spend tracking
A 6-person growth team running paid campaigns across Google Ads, Meta, and LinkedIn needs daily spend and conversion data in one place without building custom API integrations. Windsor.ai is the right call here: the Query Connector Data tool pulls blended metrics from all sources using a single 'all' connector query, and the List Connector Fields tool shows exactly which metrics each platform exposes. The API_KEY auth means setup takes under 10 minutes. The trade-off boundary: if you need real-time data (sub-hourly refresh), Windsor.ai's batch sync model will lag behind direct API calls. But for daily standup reviews and weekly budget reallocation, this MCP turns a 3-day engineering project into a 20-minute Switchy workspace setup.
When this MCP scales cross-client dashboard prep
A 12-person agency managing 40+ client ad accounts needs to pull performance data into templated reports every Monday morning. Windsor.ai's List Co-User Linked Accounts tool is built for exactly this: it enumerates all client connections without exposing tokens, and the Query Connector Data tool can loop through accounts programmatically. The 6-tool catalog is lean enough that junior account managers can learn the discovery flow (List Connectors → List Connector Fields → Query) in one training session. The threshold: if clients demand custom metrics or calculated fields beyond what Windsor.ai's connectors expose, you'll hit the schema ceiling fast. But for standard ROAS, CPC, and conversion reporting across Facebook, Google, and TikTok, this MCP eliminates the weekly manual CSV export ritual and turns it into a 5-minute Switchy prompt.
When Windsor.ai clarifies attribution for product teams
A 4-person e-commerce product team running Shopify, Google Shopping, and Amazon ads needs to understand which channel drove last week's revenue spike during a feature launch. The Query Connector Data tool pulls transaction-level data from all three sources, and the List Connector Options tool reveals platform-specific filters (date range, campaign ID, product SKU) without reading vendor docs. The single API_KEY auth means the ops lead can connect all sources in under an hour. The caveat: Windsor.ai normalizes schema across connectors, so if you need Amazon's proprietary attribution model or Shopify's custom checkout fields, you'll need to supplement with direct API calls. For high-level channel comparison and weekly product-marketing sync, though, this MCP answers 'which channel won' in 2 minutes instead of 2 hours of spreadsheet archaeology.
Frequently asked
What does the Windsor.ai MCP do in Switchy?
It pulls marketing attribution data from 200+ ad platforms into your Switchy workspace. You can query spend, conversions, and campaign performance across Google Ads, Meta, TikTok, and others without writing API code. The MCP surfaces Windsor.ai's unified schema so your team can compare cross-platform metrics in one place.
Do I need a Windsor.ai account to use this MCP?
Yes. You authenticate with a Windsor.ai API key, which means you need an active Windsor.ai subscription. The MCP won't create connectors for you—it reads from sources you've already linked inside Windsor.ai's dashboard. If your team hasn't connected ad accounts there, the MCP returns empty data.
Can the Windsor.ai MCP write data back to ad platforms?
Yes, but only for platforms Windsor.ai supports write actions on. Use the List Connector Actions tool to see which connectors expose write endpoints and what parameters they require. Most connectors are read-only; write support is limited to a subset like Google Ads campaign updates.
How is this different from querying Windsor.ai's API directly?
The MCP wraps Windsor.ai's REST API so your AI agent can discover connectors, fields, and options without you hardcoding them. If you're already comfortable with Windsor.ai's API docs and want full control, skip the MCP. If you want the AI to explore available data sources dynamically, use the MCP.
Who on the team should connect the Windsor.ai MCP?
Whoever holds the Windsor.ai API key with read access to your connected ad accounts. That's usually a marketing ops lead or the person who manages your attribution stack. The API key isn't scoped per-user, so one connection covers the entire team's Switchy workspace.