Content
73%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 well-structured, clearly sequenced planning workflow with explicit per-step completion checkpoints and a closing evidence-interpretation loop. Its main weakness is actionability: much of the guidance frames how to think rather than giving concrete, executable steps, and a few sentences are philosophically padded.
Suggestions
Tighten abstract framing sentences (e.g., the "The purpose is not to select a familiar technique..." passage) into direct imperatives that tell Claude what to produce.
Add one or two concrete worked mini-examples of an evidence path for a representative change type so the abstract guidance becomes executable.
Reposition the non-numbered "Set a verification budget" and "Scope" sections so the six-step sequence reads as one uninterrupted flow, or fold them into the relevant numbered steps.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly lean and assumes Claude's intelligence without explaining basic concepts, but contains a few philosophical sentences (e.g., "The purpose is not to select a familiar technique. The purpose is to decide how this system can reveal the truth...") that could be trimmed. | 4 / 5 |
Actionability | Concrete directives exist ("ask @librarian for focused research", "assign one owner", "choose the minimum non-duplicative evidence") alongside clear "Complete when" criteria, but much of the guidance is abstract framing ("Derive possible evidence paths from the system itself: its controllable inputs, observable effects...") rather than executable steps. | 3 / 5 |
Workflow Clarity | A clear six-step numbered sequence (Frame, Design, Affordance, Research, Runnable, Close) where every step has an explicit "Complete when" checkpoint and the Close step reports established/limited/refuted, providing explicit validation gates and a feedback loop; the destructive/batch cap does not apply because the skill plans verification rather than performing destructive operations. | 5 / 5 |
Progressive Disclosure | A single self-contained file with well-organized ## headers and numbered steps, no nested or buried references and no content that clearly belongs in separate files; scored 4 rather than 5 because it exceeds the under-50-line simple-skill threshold and a couple of sub-sections ("Set a verification budget", "Scope") sit slightly awkwardly between numbered steps. | 4 / 5 |
Total | 16 / 20 Passed |