Content
92%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 instruction-only skill body: concrete templates, explicit do/don't rules, a pre-send validation heuristic, and a loop cutoff, all in a tight 44 lines. The only nit is slight redundancy across the low-information acknowledgment examples.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean and substantive throughout with no concept over-explanation; minor trimming possible in the acknowledgment examples ("Got it", "Thanks", "ok", "done", "收到", "好的" overlap in intent). | 4 / 5 |
Actionability | Copy-paste-ready message templates ("@coordinator:domain TASK_COMPLETED: <summary>"), concrete do/don't lists, and an executable pre-send filter ("remove all Matrix IDs… send only if the remaining text contains a concrete completion, blocker, question, requested answer, or decision"). | 5 / 5 |
Workflow Clarity | The pre-send filter is an explicit validation checkpoint and the two-round loop safeguard bounds the process; as a simple single-purpose skill, the decision flow is unambiguous with no validation gaps. | 5 / 5 |
Progressive Disclosure | Under 50 lines with no need for external references; content is fully inlined with clean, well-organized sections (Routing, Coordinator Identity, @Mention Rules, History Context, Loop Safeguard) that make navigation trivial. | 5 / 5 |
Total | 19 / 20 Passed |