agentleFS
kelpi-aikelpi-ai/meta-ads-skills/skills/budget-pacer/SKILL.md

budget-pacer

Skill3 starsChanged 5 months agoMITView on GitHub

Propose budget rules and schedules (spend caps, loser-pauses, winner scaling in 20% steps) as drafts you approve. Use once an angle has proven itself and it is time to scale without breaking it.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If the loop math says the winner still is not profitable at scale, say that instead of scaling it.

    line 34

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Budget Pacer

Doctrine

"Every time I touch a winning campaign it crashes" is a scaling story, not a curse: big budget jumps reset learning, and panic edits kill compounding winners. Pacing means the boring version: scale in ~20% steps, no more than once a day, pause what has proven it loses, and let rules carry the discipline your fingers do not have. The loop decides the ceiling: scale only what returns its spend fast enough to fund the next round.

When to use

  • An angle has spent 2-3x target cost-per-result and holds. Time to feed it.
  • You keep manually raising and lowering budgets on vibes.
  • Needs a Meta Ads MCP with write access; everything is proposed as a draft first.

Run it

Budget pacing proposal, draft only, change nothing yet:

1. Read the last 14 days by ad set: spend, results, cost per result.
2. Identify what has EARNED a raise (enough spend to judge, cost per result at or under my target of [$X]) and what has earned a pause (2-3x target spent, no results).
3. Propose:
   - a pause rule: pause any ad set whose cost per result exceeds [$X * 1.5] over a 3-day window,
   - a scaling plan for the winner: +20% budget steps, at most one step per day, with the next review date,
   - a spend guardrail: a rule that alerts me if daily account spend exceeds [$CAP].
4. Show me the rules exactly as they would be created. I will say "create" for the ones I want.

Guardrails

  • Drafts until the human says "create". Spend changes are the most sensitive write there is.
  • +20% steps, not doublings. A budget doubling is a learning reset wearing a growth costume.
  • Pause rules use my numbers ([$X] comes from my loop math), never an invented benchmark.
  • If the loop math says the winner still is not profitable at scale, say that instead of scaling it.

Good output looks like

Two or three rules you can read and approve by name, a winner scaling schedule with dates, and no dollar moved without a yes.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.