Content
86%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A lean, well-structured instruction-only skill with concrete tool calls, a clear sequenced workflow, and explicit validation checkpoints. Its main gaps are the absence of a worked projection template and a fallback path when account history is missing.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Every line instructs rather than explains — no concept tutorials, no padding, and it assumes competence ('a split without projected outcomes is a guess'), matching the 'every token earns its place' anchor. | 5 / 5 |
Actionability | Concrete tool names (`list_connected_accounts`, `pull_*_ads_performance`, `forecast_campaign`) and concrete outputs (CPM/CPC/CPA/reach, two-week test flight) make the guidance mostly executable, but 'Show the math' offers no template or worked example of the projection format. | 4 / 5 |
Workflow Clarity | A clear 4-step sequence with explicit checkpoints ('Confirm what is available first', 'a clear read before wider spend', 'wait for approval'), but there is no error-recovery path for when account history is unavailable or a forecast disagrees with the target CPA. | 4 / 5 |
Progressive Disclosure | At ~33 lines with well-organized sections and no detail that belongs in a separate file, the body satisfies the under-50-line exception; sibling-skill pointers (**conversion-tracking**, **launch**) are clearly signaled and one level deep. | 5 / 5 |
Total | 18 / 20 Passed |