google-ads
Query, audit, and optimize Google Ads campaigns. Supports two modes: (1) API mode for bulk operations with the google-ads Python SDK, (2) attached-browser mode for users without API access. Use when asked to check ad performance, pause campaigns or keywords, find wasted spend, audit conversion tracking, or optimize Google Ads accounts.
86
86%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Advisory
Suggest reviewing before use
Quality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is an excellent skill description that clearly communicates specific capabilities, provides two operational modes for different user contexts, and includes a comprehensive 'Use when' clause with natural trigger terms. It is well-structured, concise, and highly distinguishable from other skills.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: query, audit, optimize campaigns, bulk operations via API, pause campaigns/keywords, find wasted spend, audit conversion tracking. Also specifies two distinct operational modes. | 3 / 3 |
Completeness | Clearly answers both 'what' (query, audit, optimize Google Ads campaigns with two modes) and 'when' with an explicit 'Use when...' clause listing specific trigger scenarios like checking ad performance, pausing campaigns, finding wasted spend, etc. | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural terms users would say: 'ad performance', 'pause campaigns', 'keywords', 'wasted spend', 'conversion tracking', 'optimize', 'Google Ads accounts'. These are terms users naturally use when working with Google Ads. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive with a clear niche around Google Ads specifically. The mention of google-ads Python SDK, attached-browser mode, and Google Ads-specific terminology makes it very unlikely to conflict with other skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
72%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured, actionable skill that covers two distinct usage modes with concrete code and navigation paths. Its main weaknesses are redundancy between the Safety Boundaries and Security Addendum sections, and the lack of explicit validation/confirmation checkpoints embedded within the destructive operation workflows themselves (despite mentioning the requirement in safety sections). The empty Troubleshooting header is a minor oversight.
Suggestions
Merge the 'Safety Boundaries' and 'Security & Change-Control Addendum' sections into a single concise section to eliminate redundancy.
Add explicit confirmation and validation steps directly within the pause/mutate workflow code blocks (e.g., print impacted entities, prompt for confirmation, verify post-mutation status), rather than relying solely on a separate safety section.
Remove or populate the empty '## Troubleshooting' header — currently it appears before the Safety Boundaries section with no content.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Generally efficient but has some redundancy — the Safety Boundaries and Security & Change-Control Addendum sections overlap significantly (both mention requiring confirmation before mutations, protecting credentials). The Troubleshooting header is empty. Some explanatory text like 'Quick health check for any Google Ads account' is unnecessary filler. | 2 / 3 |
Actionability | Provides executable Python code for API mode with real GAQL queries, specific browser navigation paths, concrete filter instructions, and copy-paste ready code for mutations. The audit checklist gives specific thresholds and paths. Both modes have concrete, actionable workflows. | 3 / 3 |
Workflow Clarity | Steps are listed clearly for both modes, but mutating operations (pausing keywords via API) lack explicit validation/confirmation checkpoints in the workflow itself. The safety boundaries mention requiring confirmation but the actual pause workflow code doesn't include a verify-before-execute step or error handling/rollback guidance. For destructive batch operations like pausing keywords, this gaps caps the score at 2. | 2 / 3 |
Progressive Disclosure | Well-structured with clear sections for mode selection, browser mode, API mode, audit checklist, and output formats. References to external files (browser-workflows.md, api-setup.md) are one level deep and clearly signaled. Content is appropriately split between the overview skill and reference files. | 3 / 3 |
Total | 10 / 12 Passed |
Validation
72%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 8 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
metadata_version | 'metadata.version' is missing | Warning |
metadata_field | 'metadata' should map string keys to string values | Warning |
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 8 / 11 Passed | |
- Repository
- jdrhyne/agent-skills
- Commit
6768672
- Reviewed
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.