Campaign Health Check

Decide which campaigns to scale, fix or stop

Raised by customers on 100+ calls in the past year

Ranks live campaigns by positive replies and meetings, not opens, flags senders close to bounce limits, and suggests the one change for each.

Reporting

Luna

Reviewed by Alta

What it does

Reads every live campaign
Ranks them by positive replies and meetings, not opens
Flags senders close to bounce or complaint limits
Suggests the one change for each: scale, fix or stop

You give it

Which campaigns to include
How often to run it

You get

Every live campaign ranked: scale, fix or stop
Sender risk flags
One suggested change per campaign

Why it works

Opens and clicks hide what matters. Ranking by replies and meetings shows which campaigns earn more volume and which need a change.

Run this play

Copy the SKILL.md below into your agent

Works in Alta, Claude, Cursor and Codex

Runs on

Luna

Category

Reporting

CostAlta credits per run, shown before it runs

ApprovalEvery run stops at a draft for you

--- name: campaign-health-check description: Rank live campaigns by positive replies and meetings, flag sender risk, and suggest one change for each. agent: Luna category: Reporting ends_in: report --- # Campaign Health Check A weekly look at what to scale, what to fix and what to stop. ## How it runs 1. **Read every live campaign** 2. **Rank by positive replies and meetings**, not opens 3. **Flag senders** close to bounce or complaint limits 4. **Suggest the one change** most likely to help each campaign 5. **Write the report** - scale, fix or stop ## Rules - Rates use delivered messages as the base - Nothing is changed - the play only suggests

- Campaigns are ranked by positive replies and meetings, with rates on delivered messages - Each campaign gets exactly one suggested change, not a list - Sender risk is flagged before it becomes a deliverability problem

The play fileSKILL.md

This is the whole play. Read it, copy it, or change it for your team - it runs the same in Alta, Claude, Cursor and Codex.

Install this play

Claude app or claude.aiUpload the zip as a skill

Download the zip, then upload it as a skill in your Claude settings. Code execution needs to be on.

Claude CodeDrop the folder in

Unzip it into your skills folder. Claude Code finds it on its own.

~/.claude/skills/

Cursor and other agentsAdd the SKILL.md

Put the folder where your agent reads skills, for example:

.cursor/skills/

Run it where you work

Same play, same result. Pick the place you already are.

In Alta

One click

Open Skills in Alta, pick the play and press Run. Or ask Ask Alta to run it for the companies you care about.

In Claude

Ask in plain words

Connect Alta to Claude and describe the target. Claude loads the play, the agent does the work, and the draft opens in Alta for you to approve.

In Cursor or Codex

Drop in the file

Add the SKILL.md to your agent's skills folder with the Alta plugin connected. Edit it like code, and keep your team's version in the repo.