Content
61%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 coherent, well-structured prompt template with a complete output schema and a useful protocol taxonomy, but it under-specifies its own decision logic: the connection-protocol family has no selection rule, and the hard JSON-validity constraint has no validation checkpoint. A redundant bilingual notes section duplicates earlier constraints and pushes the file past lean-skill size. These are fixable gaps in an otherwise sound single-file skill.
Suggestions
Add a selection rule for the third protocol family (e.g., 'If the context involves other people or social energy, design a connection protocol'), and a final checkpoint like 'Verify output is valid JSON matching the schema before responding' — this addresses the two workflow_clarity gaps.
Delete or merge the 注意事项 section, since three of its four bullets restate the Constraint section verbatim — this trims redundancy and tightens conciseness.
Include one worked example of a completed output JSON (e.g., a filled NSDR Reset protocol) to anchor the schema and push actionability toward copy-paste readiness.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly efficient — role, task, constraints, input context, output schema, and a compact protocol taxonomy, with no padding explaining things Claude already knows. However, the 注意事项 section largely restates constraints already given ('所有建议必须有明确的科学依据' duplicates 'Strict Grounding', '避免模糊的建议' duplicates 'Avoid generic advice'), which is unnecessary explanation that could be trimmed. This fits 'Mostly efficient but includes some unnecessary explanation or could be tightened', not 4 because the redundancy is a real (if small) block rather than a minor instance. | 3 / 5 |
Actionability | As an instruction-only skill it gives concrete, specific guidance: a complete JSON output schema with field-level examples, an explicit micro-step format ('Do X for Y mins'), strict grounding rules, and a concrete taxonomy of protocol families (NSDR, 冥想, 深度工作启动仪式...). This matches 'Mostly executable guidance; concrete code or commands with minor gaps' — the gap being no worked example of a filled output JSON and no specified step count/duration calibration, so it does not reach 5's copy-paste-ready, common-cases-covered level. | 4 / 5 |
Workflow Clarity | The structure implies a rough sequence (read [CONTEXT] → apply philosophy → emit JSON), but the core decision logic is incomplete: the Philosophy section maps only two branches (context suggests rest → recovery; context suggests focus → engagement) while the taxonomy lists a third family (连接型/connection) with no rule for when to choose it. There is also no checkpoint validating the output (e.g., 'verify JSON fits schema before responding'), even though valid JSON output is a hard constraint. This matches 'Steps listed but validation gaps; sequence present but checkpoints missing or implicit' — not 4, because the missing connection-protocol branch and absent output validation are substantive gaps rather than minor ones. No destructive/batch cap applies. | 3 / 5 |
Progressive Disclosure | No bundle files exist, and this self-contained prompt template has no need for external references; its well-labeled sections (Role, Task, Constraint, Input Context, Output Schema, Philosophy, typed protocol sections) make navigation easy. The body is ~63 lines — just past the 'under 50 lines' threshold where well-organized sections alone earn a 5 — and the inlined protocol taxonomy plus the duplicated constraint section could be flattened or split. This fits 'Good structure; most content is appropriately placed; references mostly clear; minor organization gaps' rather than the clear-overview-with-signaled-references anchor. | 4 / 5 |
Total | 14 / 20 Passed |