Content
75%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 content is a dense, well-sequenced instruction skill that assumes Claude's competence and delivers specific, actionable guidance throughout. It is held back only slightly by minor prose that could tighten and a few unspecified placeholders.
Suggestions
Tighten the narrative asides in Step 4 and Step 5 (e.g., 'The highest-leverage move in a portfolio is...') to pure directives to lift conciseness.
Replace unspecified placeholders like '>$Xm in AP volume' with a concrete threshold or an explicit instruction to let the user set it.
Add an explicit data-completeness check before Step 3 ranking (e.g., confirm each company has sector, revenue, headcount, tech stack) to add a validation checkpoint to the workflow.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes domain competence (uses EBITDA, CIM, LTM, prior-auth, MSA without defining them), with every section delivering actionable guidance; a few narrative asides in Steps 4–5 could be trimmed slightly. | 4 / 5 |
Actionability | Gives concrete, specific guidance — exact fields to extract, three gate questions, named leverage-point patterns, three ranking dimensions, and a filled output table template — though it leaves a few placeholders ('$Xm in AP volume') unspecified. | 4 / 5 |
Workflow Clarity | A clearly sequenced five-step workflow with decision checkpoints (the gate questions in Step 2, the Go/Wait gates, the 'What we're NOT doing' filter); minor validation gaps around confirming data completeness before ranking. | 4 / 5 |
Progressive Disclosure | Well-organized into labeled sections (Workflow, Important Notes) with no nested references and no bundle files present; the ~95-line self-contained body is appropriately structured, though a couple of reference-style lists could be split out. | 4 / 5 |
Total | 16 / 20 Passed |