Content
82%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.
The body is a well-organized, highly actionable overview that correctly delegates deep material to seven real reference files and includes concrete templates, numbers, and checklists. Weaker spots are mild redundancy in the question-gathering sections, a missing validation thread through the design workflow, and two second-level reference chains that slightly complicate navigation.
Suggestions
Merge "Before Starting" context questions and "Task-Specific Questions" into one section — they ask nearly the same things (program type, LTV/CAC, tools, budget, shareability) and duplicate each other.
Add explicit validation checkpoints to the program-design workflow (e.g., verify tracking and attribution data before launch, review funnel metrics before scaling incentives), extending the pre/post-launch checklist discipline into the design steps.
Flatten the two-level reference chains by linking review-program-operations.md directly from SKILL.md's Review Programs section and consolidating the six repeated viral-mechanisms.md pointers into one clearly labeled callout.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense and free of basic-concept padding (ranked lists, tables, checklists, templates), but "Before Starting" context questions and "Task-Specific Questions" substantially duplicate each other (program type, LTV/CAC, tools, budget, shareability), and viral-mechanisms.md is pointed to six times — minor trimming opportunities that keep it below the every-token-earns-its-place anchor. | 4 / 5 |
Actionability | Concrete, ready-to-use guidance throughout: "say '$10 off,' not '40% off,'" a copy-paste email template with subject line, a problem-to-fix table, ranked share mechanisms, and specific A/B test lists — specific examples cover the common cases for this instruction-only skill, matching the score-5 anchor rather than the minor-gaps anchor. | 5 / 5 |
Workflow Clarity | Clear sequences exist (Step 1-3 design flow; Before Launch → Launch → Post-Launch checklists) with some checkpoints ("Test complete referral flow," "Review conversion funnel," "Fix friction points"), but the design workflow has no verification gates and the design → launch → optimize progression is implied rather than one coherent validated workflow — clear sequence with minor validation gaps fits the score-4 anchor. | 4 / 5 |
Progressive Disclosure | References are well-signaled and content-labeled ("For examples and incentive sizing: See references/program-examples.md"), and all referenced files exist, but chains run two levels deep (SKILL.md → review-programs.md → review-program-operations.md; viral-mechanisms.md → program-examples.md), which breaks the strict one-level-deep requirement of the score-5 anchor — good structure with minor organization gaps fits score 4. | 4 / 5 |
Total | 17 / 20 Passed |