Content
26%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 token-efficient but vacuous: it provides no executable guidance, no workflow, and no structure. Its only strength is brevity, while every substantive dimension fails.
Suggestions
Replace 'Do the task. Use your judgment.' with concrete, executable instructions (specific commands or code covering the common cases).
Add clear section headers (e.g., a Quick start) so the single task is unambiguous, which would also raise progressive_disclosure.
If the task involves destructive or batch operations, include explicit validation/verification steps so the workflow can score above 3.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is two short sentences with no padding and assumes Claude's competence, but the tokens convey no useful information so they fall just short of fully earning their place, sitting between the efficient (4) and lean (5) anchors. | 4 / 5 |
Actionability | 'Do the task. Use your judgment.' gives no concrete code, commands, or specific steps, matching the anchor for entirely vague, abstract direction that only describes rather than instructs. | 1 / 5 |
Workflow Clarity | There is no sequence, no defined steps, and no validation; the single instruction 'Do the task' is maximally ambiguous, so the simple-skill exception does not apply and it matches the score-1 anchor. | 1 / 5 |
Progressive Disclosure | There are no section headers or organization of any kind, only a title and two sentences, giving minimal structure; it is not a monolithic wall of text, so it sits above score 1 but below the 'some structure' of score 3. | 2 / 5 |
Total | 8 / 20 Passed |