Skip to content
SkillModel-led349 tokens in use

ads-validate

Validates Claude Ads contracts, scores, run bundles, capabilities, installs and release readiness, and demotes readiness when evidence is stale.

11,191 stars on the repository MIT LicenseAgriciDaniel/claude-ads · skills/ads-validate
/plugin marketplace add agricidaniel/claude-ads /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-validate costs

66tokens idle
349tokens 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 →
Fair
  • Frontmatter8/15
  • Structure6/25
  • 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-validate picks the narrowest validation target and runs it: a contract or bundle through the deterministic core validator, scores, a run’s manifest and source lineage, a capability, the whole repository, or a release through a fresh-context verifier.

Results come back as machine-readable pass or fail with the highest-priority blocker, exact evidence and recovery steps. It also covers install and uninstall safety, allowing removal only of paths listed in the ownership manifest and rejecting pipe-to-shell install instructions.

What You Get

  • A pass or fail result in machine-readable form for the chosen target
  • The highest-priority blocker with exact evidence
  • Recovery steps
  • A failed current-evidence check when an expired claim cannot be reverified
  • An install and uninstall safety verdict

Common Use Cases

  • Checking a run bundle before rendering a report
  • Verifying that a platform capability has tests and sources behind it
  • Confirming an uninstall will only remove files the toolkit owns
  • Running release readiness gates before tagging a version

What Makes It Unique

Stale evidence and skipped remote CI demote readiness. An uninstall may remove only exact paths in the matching ownership manifest, the whole manifest is validated before any deletion, and an ads-* glob is never used, so an unrelated skill such as ads-weather stays untouched. Pipe-to-shell install instructions are rejected, and archives need a SHA-256 checksum verified against a trusted release channel.

Under The Hood

The skill is written instructions with no scripts of its own. The agent calls the Claude Ads core validator and repository gates, which run commands on your machine, and dispatches a fresh-context verifier subagent for releases. It reads manifests and run files and can check install and uninstall paths, and it stops before any deletion if the ownership manifest is absent, invalid, mismatched or unsafe. Expired claims are handed to ads-research for reverify-or-demote.

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