IonBin
All pages
Search

Write a weekly ad report from connected account data

Data · Content · AdsChecked 2026-10-17

Prepare a reconciled weekly summary for human review.

  1. 01Windsor.ai
  2. 02Claude
  3. 03You
Recipe context

Connect Windsor.ai’s MCP server to Claude, ask for last week’s spend and results by channel, and edit the summary before it goes to the team.

Independent review. Unpaid; no partnership with the tools on this page. Based on public documentation and research. Trademarks identify their owners’ products. Disclosure.

Editorial shortlist

Compare your options.

For reconciling campaign metrics for weekly reporting.

Windsor access is separate; align refresh and attribution settings. Choose one for analysis of a reviewed data extract; switching may need setup changes.

Used by Ion · separate from rank.

Unpaid picks. No partnership with ChatGPT, Julius AI, OpenAI Codex.

  1. 1 · Primary / high

    ChatGPT

    Why it fits: File analysis can join calculation review with a narrative explanation.

    Tradeoff: An answer is not a reconciled report; verify calculations and upload limits.

    Prefer it when the reviewer needs both analysis and an explanation for colleagues.

    Evidence for ChatGPT

    The linked IonBin review records capabilities, access limits and source links. Review date: 2026-10-18. Primary product reference ↗

  2. 2 · Secondary / medium

    Julius AI

    Why it fits: A data-analysis workspace keeps the supplied table and business question together.

    Tradeoff: Wrong joins, missing rows and inconsistent units can produce convincing but incorrect charts.

    Choose it for interactive exploration of a cleaned extract.

    Evidence for Julius AI

    The linked IonBin review records capabilities, access limits and source links. Review date: 2026-10-18. Primary product reference ↗

  3. 3 · Third / lower

    OpenAI Codex

    Used by Ion — Used by Ion for its own work. This is separate from recommendation rank and is not a claim that every feature was tested.

    Why it fits: A coding workspace keeps calculations and validation checks in reviewable files.

    Tradeoff: Someone must read, run and maintain the code and protect sensitive files.

    Prefer it when repeatability and code review matter more than conversational convenience.

    Evidence for OpenAI Codex

    The linked IonBin review records capabilities, access limits and source links. Review date: 2026-10-18. Primary product reference ↗

How this shortlist was chosen

Editorial judgment, not a personal-use endorsement or performance score. These are alternatives for the named role, not three required stack components. Vendors did not pay for, request or approve these positions.

Inspectable calculations come before narrative convenience. A lower-ranked option may fit your constraints better; compare the conditions stated on each card before choosing.

Setup and portability: Record export date, filters and units; save calculations with the result. The source ledger still needs reconciliation.

Reviewed 2026-10-21. This page applies a shared comparison for “analysis of a reviewed data extract” to the stated task and constraints. It is documentation-based editorial guidance, not three new hands-on tests. We weigh value for the job, distinct capabilities, replacement limits, technical effort and interoperability. Check current plan terms before adopting or switching.

See our review method, MCP connections and portability and switching costs.

Connects through: MCP

Steps

  1. Windsor.ai

    Supplies cross-channel ad data through get_data on the Windsor.ai MCP server

    The same server has write actions that can pause campaigns or change budgets; keep approval on.

    MCPConfirmedSource ↗
  2. Claude

    Writes the summary and flags changes

    AI platformConfirmed
  3. You

    Check figures against the ad platforms and approve

    Human stepConfirmed

Confirmed = vendor-documented, not Ion-tested. Human and AI-platform steps are proposed roles.

Human approval

You approve the report.

Watch out for

Windsor.ai's MCP server can also pause campaigns and change budgets. Keep tool approval on and decline any write call; this workflow only reads.

Starter prompt

Pull last week's and the prior week's spend, clicks, conversions and cost per conversion by channel from Windsor.ai. Write a summary of no more than 150 words and flag changes over 20%. Do not call any write action. I will check the totals against each ad platform and approve the report.

Written for this workflow. Paste it into your AI platform and adjust it to your situation.

Tools in this workflow

ToolRoleOfficial site
Windsor.aiMCP windsor.ai ↗
ClaudeAI platform claude.ai ↗

Which AI platforms can run it

PlatformStatusWhy
Claude (incl. Claude Code) To verify Windsor.ai: MCP pairing needs route-specific evidence; a named plugin or CLI does not establish it
OpenAI Codex To verify Windsor.ai: MCP pairing needs route-specific evidence; a named plugin or CLI does not establish it
Google Antigravity To verify Windsor.ai: MCP pairing needs route-specific evidence; a named plugin or CLI does not establish it
ChatGPT To verify Windsor.ai: MCP pairing needs route-specific evidence; a named plugin or CLI does not establish it
Gemini To verify Windsor.ai: MCP pairing needs route-specific evidence; a named plugin or CLI does not establish it
Perplexity To verify Windsor.ai: MCP pairing needs route-specific evidence; a named plugin or CLI does not establish it
What the platform statuses mean

Confirmed: documented access meets the recorded requirements. Inferred: a reviewed note describes a plausible, untested pairing. To verify: a connection or step is unresolved. MCP and API pairings need route-specific evidence; a plugin, CLI or API alone does not prove compatible access, credentials or permissions. These are documentation assessments, not IonBin runtime tests.

What we checked (2026-10-17)

  • Windsor.ai MCP (mcp.windsor.ai) signs in with OAuth, or with an API key if your client only supports keys. get_data runs a query by date range, account and fields, and get_connectors lists what you’ve connected.
  • Write actions. The same server can run write actions through execute_action, such as pausing or enabling campaigns and setting budgets on Meta Ads, Google Ads and others. For a report, keep tool approval on and decline any write call.
  • Scope and cost. Available fields, refresh timing and access depend on the connected source and plan. Check current pricing and account limits before scheduling reports.

Set it up

  1. Connect your ad accounts in Windsor.ai, then connect the MCP server to Claude.
  2. Ask: “Pull last week’s and the prior week’s spend, clicks, conversions and cost per conversion by channel. Write a summary of no more than 150 words and flag changes over 20%. Don’t call any write actions.”
  3. Check the totals against each ad platform’s own dashboard. Check reporting currency, time zones, attribution windows, conversion definitions and reporting lag. Cross-platform conversions can overlap; do not add them as unique customers.
  4. Record extraction time and any incomplete data. Edit the summary and send it yourself.

This is a proposed reporting sequence supported by documented tools, not a tested account integration. Restrict available tools where the client permits it; a prompt saying not to write is not an authorization control.

If you connect Google’s own Ads MCP server instead of a connector, the Google Ads Account Diagnostics skill adds structured checks for conversion drops and lost impression share. IonBin reviewed it from source only and didn’t run it.

Sources

Related workflows

No affiliate links on this page. No paid placements in workflows. How IonBin makes money.