Content
85%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 tight, well-structured principle skill: it states the rule, gives concrete do/don't guidance with a precise ID taxonomy, and disambiguates with a BAD/GOOD example. Only minor gaps — an illustrative rather than executable code example and a low-value provenance section — keep it from full marks.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean with zero over-explanation of known concepts ('Never rely on "latest" or "current" when crossing process or async boundaries', terse DO/DON'T bullets). The only minor trimmable content is the 'Source Sessions' provenance block, which adds tokens that do not instruct, keeping it just below anchor 5's 'every token earns its place'. | 4 / 5 |
Actionability | Concrete, specific guidance throughout: 'Pass `--session-id $ID` when spawning processes', 'Use full UUIDs, not partial matches', a precise ID taxonomy (session_id / root_span_id / turn_span_id), and a BAD/GOOD code contrast. The example uses illustrative `spawn('analyzer', ...)` pseudocode rather than an executable API, so it falls short of anchor 5's copy-paste-ready bar. | 4 / 5 |
Workflow Clarity | This is a simple, single-purpose behavioral rule and the single action is unambiguous: never query 'most recent' at execution time, always pass explicit IDs across boundaries, and keep ID types separate. The DO/DON'T lists plus the worked example leave no ambiguity about what to do, and no destructive/batch-operation validation cap applies. | 5 / 5 |
Progressive Disclosure | The skill is under 50 lines, needs no external references, and is cleanly organized into Pattern / DO / DON'T / Example / Source Sessions sections. Per the simple-skill guideline, well-organized sections with no bundle dependencies merit the top score. | 5 / 5 |
Total | 18 / 20 Passed |