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.
An exemplary lean instruction skill: dense, specific, and free of padding, with a clear step sequence and built-in checks (stock confirmation, send-limit totals, blackout flags). The main improvements are structural — unpacking the run-on Reads line and spelling out error-recovery loops for unconfirmed stock or exceeded send limits.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is a lean Reads/Does/Writes structure where every clause carries operational content — exact paths ("lifecycle/calendar/<YYYY-MM>.md"), specific fields ("subject + 2 alternates, preview, body, CTA with tracked URL, plain text"), and concrete rules ("60–70% useful sends in non-sale weeks"). Nothing explains concepts Claude already knows; every token earns its place, matching the score-5 anchor. | 5 / 5 |
Actionability | As an instruction-only skill it gives highly concrete output specs — per-send field checklists, exclusion rules, exact write paths, test-winner criteria — so the absence of code is not penalized. It stops short of fully executable because several items defer to unshown context ("offer if any (from the ladder only)", "Margin box from offer-math", "the one idea") whose formats are defined elsewhere, leaving minor gaps. | 4 / 5 |
Workflow Clarity | The four numbered steps give a clear sequence (calendar, then this week's finished sends, then SMS, then margin boxes) with most checkpoints present: "stock confirmed for each featured product", "Totals the month's sends ... against the plan's monthly send limit; spreads the sends or proposes the upgrade", and "Flags blackout dates and shipping cut-offs". It lacks the explicit validate-fix-retry feedback loops of the score-5 anchor, e.g. what to do when stock is unconfirmed or the send limit is exceeded beyond 'proposes the upgrade'. | 4 / 5 |
Progressive Disclosure | The body is a compact overview that correctly delegates detail to one-level-deep references in the parent skill, each signaled with its purpose ("copy-craft.md", "sms-push-whatsapp.md", "modes-and-incidents.md (seasonal)"). It sits at 4 rather than 5 because the references are packed into a single dense run-on 'Reads' sentence alongside tool calls and data sources rather than presented as clearly navigable, individually-signposted pointers. | 4 / 5 |
Total | 17 / 20 Passed |