Skip to content
SkillModel-led287 tokens in use

ads-optimize

Diagnose a paid-ad account and draft ranked optimizations for budgets, bids, search terms and creative, with a refusal on permanent deletion and a gate on apply.

11,191 stars on the repository MIT LicenseAgriciDaniel/claude-ads · skills/ads-optimize
/plugin install claude-ads@ai-marketing-hub-claude-ads
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing ads-optimize costs

71tokens idle
287tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

Nothing detected: no shell, network, file writes, git or credentials in its files.

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Good
  • Frontmatter8/15
  • Structure14/25
  • Examples0/15
  • Output15/15
  • Size15/15
  • Resources10/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent

No findings.

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Good
  • Development activity84
  • Maintenance100
  • Community56
  • Maturity32
  • Momentum40

ads-optimize diagnoses paid-ad performance and drafts changes. It loads the latest normalized snapshot, prior decisions, monitoring results, experiment state, conversion lag, economics and the platform capability manifest, then compares no-change, experiment and mutation options.

It defaults to —draft. Applying a change goes through the main mutation gate, and permanent deletion of campaigns, ad groups, ads, audiences or conversions is refused.

What You Get

  • Ranked recommendations with confidence and success measures
  • A comparison of no-change, experiment and mutation options
  • Mutation plans for approved recommendations, produced through the main gate
  • Reversible alternatives when a delete is requested: pause, archive, label, export a backup and set a review date

Common Use Cases

  • Reallocating budget between campaigns
  • Reviewing search terms and negative keywords
  • Rotating creative and pausing weak ads
  • Improving CPA or ROAS from causal evidence and a stated success measure

What Makes It Unique

It never treats a fixed CPA multiple, budget ratio, benchmark or novelty claim as enough reason to change an account. Search-term actions need a search terms report, relevance evidence and an overblocking review, and it will not invent example negative keywords without them. A request to delete campaigns is turned into reversible options.

Under The Hood

The skill is an instruction file with no scripts of its own. The agent reads the snapshots, monitoring results and capability manifest produced by other Claude Ads steps and writes ranked recommendations. In apply mode, and only when the exact operation is enabled and the remote state still matches the draft, it asks platform adapters to change the live account, verifies the result and keeps an audit and rollback record.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers