>_

SpendSignoff documentation

The always-on operator for your ad accounts. Read and draft from your AI client — approve before anything goes live.

Knowledge base

General FAQ

Short answers to the questions buyers ask first — what it connects to, who can spend, what is live, and where the audit log lives.

Your AI can read and draft — it can never spend without your approval.

Can my AI client spend money on its own?

No. The operator can read and draft, never spend. Every live change passes the two-step Approve & push live → Confirm control inside the SpendSignoff app and lands in the audit log. The MCP server requests mcp.read and mcp.draft only — there is no mcp.approve scope.

Which ad platforms and AI clients work today?

Ad platforms live now: Google Ads and Meta. Coming soon: TikTok, Microsoft, LinkedIn, and Amazon. AI clients live now: Claude, ChatGPT, and Cursor. Coming soon: Claude Code, Codex, Windsurf, and Gemini CLI.

What does connecting an account actually grant?

The first OAuth authorization is read-only. SpendSignoff reads spend, structure, and performance so the operator can find money. Linking does not grant the ability to spend — that is gated behind your explicit approval, not an OAuth scope.

Quick answers

  • Is reading free? Yes. Reads are ungated on every plan, including Free.
  • Is there an auto-pilot? No. V1 is propose-only — the loop drafts, you approve.
  • Can I undo an approved change? Yes. One-click rollback from the KMS-signed audit log.
  • Where is the MCP server? https://mcp.spendsignoff.com/mcp over Streamable HTTP + OAuth 2.1.
  • What do the plans cost on? Connected ad accounts, not messages — flat Free / Solo / Pro / Agency.

Try it before you pay

Connect an account read-only on Free and watch the operator draft real changes. You see exactly what it would do before any plan or approval is involved.

Next

Compatibility

What each AI client supports today: Claude, ChatGPT, Cursor.

    General FAQ — SpendSignoff Docs