ads-math — Claude Code skill: paid-media math and forecasts
Works out CPA, CPL, CPC, CPM, ROAS, MER, break-even, margin and LTV:CAC. Useful when a budget or forecast needs numbers behind it.
npx skills add AgriciDaniel/claude-ads --skill ads-math -a claude-codeWhat ads-math does
The skill runs a calculation in steps. It starts by pinning down which decision you are making, then gathers units, currency, period, tax and refund handling, margin, attribution basis and how uncertain the inputs are. Every figure gets tied back to an operator value or a cited artifact, and the formula is shown.
Before any result appears, it checks denominators, signs, gaps, mismatched windows and unit conversions. Where uncertainty changes the answer, you get a base case plus downside and upside. Platform-attributed revenue, blended business revenue, cash flow and contribution margin stay in separate buckets, so the bases do not get mixed.
Output is machine-readable: inputs, formulas, results, sensitivities and what the numbers imply for the decision. Missing financial inputs are not invented, and division by zero is not papered over. A single-point forecast without its assumptions is not something this skill returns, so come ready to state those assumptions yourself.
What it helps with
- calculate CPA and CPL for a campaign against targets
- find the break-even point for a new channel
- separate ROAS and MER across different revenue bases
- forecast a quarterly budget in three scenarios
- model LTV:CAC and contribution margin before launch
Example prompts for the agent
- Calculate CPA and CPL for this campaign and show the formula and assumptions
- Find the break-even point for the new channel at a 35% margin
- Forecast next quarter's budget: base, downside and upside cases
How to install ads-math
1Skills CLI (Vercel)
npx skills add AgriciDaniel/claude-ads --skill ads-math -a claude-code2Manually: put the files in place
~/.claude/skills/ads-math/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/ads-math/
1Skills CLI (Vercel)
npx skills add AgriciDaniel/claude-ads --skill ads-math -a codex2Manually: put the files in place
.agents/skills/ads-math/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add AgriciDaniel/claude-ads --skill ads-math -a cursor2Manually: put the files in place
.cursor/skills/ads-math/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install AgriciDaniel/claude-ads ads-math2Manually: put the files in place
.github/skills/ads-math/SKILL.md1Skills CLI (Vercel)
npx skills add AgriciDaniel/claude-ads --skill ads-math -a gemini-cli2Platform CLI
gemini skills install https://github.com/AgriciDaniel/claude-ads --path skills/ads-mathManually: put the files in place
.devin/skills/ads-math/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add AgriciDaniel/claude-ads --skill ads-math -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about ads-math
How do I install ads-math in Claude Code?
Run npx skills add AgriciDaniel/claude-ads --skill ads-math -a claude-code. The skill lands at ~/.claude/skills/ads-math/SKILL.md.
What does ads-math do?
It calculates and models paid-media figures: CPA, CPL, CPC, CPM, ROAS, MER, break-even targets, contribution margin, LTV:CAC, budgets, forecasts and experiment economics.
What inputs does it need?
Units, currency, period, tax and refund treatment, margin, attribution basis and an uncertainty estimate. It will not hand back a point forecast without assumptions.
Source text
SKILL.md · MIT
--- name: ads-math description: "Calculate and model paid-media CPA, CPL, CPC, CPM, ROAS, MER, break-even targets, contribution margin, LTV:CAC, impression-share opportunity, budgets, forecasts, and experiment economics. Use for PPC math, ad calculator, break-even analysis, ROAS calculator, CPA calculator, budget forecast, LTV CAC, or MER." --- # Paid Media Financial Model 1. Identify the decision and collect units, currency, period, tax/refund treatment, margin, attribution basis, and uncertainty. 2. Show the formula and map every input to an operator value or cited artifact. 3. Validate denominators, sign, missing values, incompatible windows, and unit conversions. 4. Calculate base, downside, and upside cases where uncertainty affects the decision. 5. Keep platform-attributed revenue, blended business revenue, cash flow, and contribution margin distinct. 6. Return machine-readable inputs, formulas, outputs, sensitivities, and decision implications. Never fabricate missing financial inputs, hide division-by-zero, or present a point forecast without its assumptions.
© AgriciDaniel · MIT · AgriciDaniel/claude-ads
