The playbook for the job
you already own.
A solutions page follows one role through one piece of work, start to finish: the run they kick off, the report that comes back, where every threshold in it came from, and what they can say in the meeting afterwards. Read your role, not our product catalogue.
Every run reads your live account through the Amazon Ads MCP and the Selling Partner MCP — not a spreadsheet somebody exported last Tuesday.
Paid ads, run before the standup.
The audit reads every campaign in the account, not the five that carry the spend, and comes back with the math attached.
Where the spend pooled, what stopped serving, and which campaigns to consolidate — read-only against a live account, in about twenty minutes. It displaces the category of tool that charges a percentage of spend to tell you less.
Read the paid ads playbookThree pages describe the same software. They answer different questions.
What it connects to
The MCP servers, the marketplaces, the operations your agent can call. Read it when you are working out whether your stack fits.
All featuresWhat one tool does
A single workflow, its inputs, its output contract, and the guardrails around it. Read it when you know the job and want the part that does it.
All SkillsWhat your Tuesday looks like
One role, one playbook, the artifact it returns and the reasoning behind every number in it. Read it when the question is whether this changes your week.
Paid adsA role gets a page when the Skills behind it go deep enough to judge. Where that is not true yet there is no page, because a page that lists coverage instead of showing depth is one you would scroll past anyway.
Amazon Ads MCP
The server the paid ads playbook reads through: Sponsored Products, Brands, Display and TV, Amazon DSP, and Amazon Marketing Cloud.
See the serverYour AI client
Claude, ChatGPT, Cursor, or any MCP-capable runtime your team already has open. Nothing new to learn and no second place to log in.
See the clientsThe tools underneath
Every operation a playbook can call, named and documented, so you can see exactly what a run touched and what it never had access to.
Read the referenceGuides
Where the numbers in a playbook come from.
The rules behind every run are written down and argued for: when a bid can move on its own, which data mode a question is allowed to use, and what happens when the evidence will not support an answer.
The Agent Bidding Rulebook That Prevents Bid Thrash
A bid should move only when sufficiency, an ACoS deadband, and a daily clamp all agree. The bounded rulebook an agent runs, with the thresholds.
The Data-Mode Router That Stops Bad ACoS Math
Your SQP export has no cost column. Atlas blocks ACoS there and routes to Conversion Performance Index instead, the fix for nonsense report numbers.
Human Approval for AMC Activation
Gate AMC audience activation behind human approval: size the seed, review the exact payload, then submit. 2000-user floor, 24 to 36 hour DSP lag.
Frequently asked questions
What a solutions page is.
How these pages differ from the feature and Skill pages, how a category gets one, and who ends up running the work.
What is a solutions page?
How is this different from the Skills pages?
How do you decide which roles get a page?
Does this replace the person who owns the account?
We use an agency. Where do they fit?
Can we change how a playbook behaves?
The bottom line
The work is already yours. The reading is what takes the week.
Nobody on your team is short of judgement — they are short of the hours it takes to read two hundred campaigns, every listing, every return, before they can use it. That is the part that runs while they are at the standup.