Content
71%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 well-organized and highly actionable with concrete schedules, tables, and benchmarks, but it leans on motivational prose and, more seriously, points to bundle files (two references and a script) that do not exist. Fixing the dangling references is the highest-leverage improvement.
Suggestions
Create the referenced bundle files (references/cancel-flow-playbook.md, references/dunning-guide.md, scripts/churn_impact_calculator.py) or remove the links — dangling references break progressive disclosure.
Trim motivational framing ("Churn is a revenue leak you can plug. That compounds.", the role preamble) to assume Claude's competence and earn the top conciseness anchor.
Add an explicit verification checkpoint after the calculator run (e.g., sanity-check the recovered-MRR figure against current MRR) to close the workflow_clarity feedback loop.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly efficient with dense tables and concrete schedules, but motivational framing ("Churn is a revenue leak you can plug... That compounds", "A cancel flow is not a dark pattern — it's a structured conversation") and the role preamble add tokens Claude doesn't need. | 4 / 5 |
Actionability | Strong executable guidance — numbered retry schedule (3/5/7 days), a day-by-day dunning table, reason→offer mapping tables, metric formulas with benchmarks, and red-flag thresholds — but the referenced `python3 scripts/churn_impact_calculator.py` has no corresponding script in the bundle. | 4 / 5 |
Workflow Clarity | The 5-stage cancel flow and dunning sequence are clearly sequenced with metrics/red-flags acting as checkpoints, but there is no explicit validate→fix→retry loop and the calculator invocation has no stated verification of its output. | 4 / 5 |
Progressive Disclosure | Section structure is clean and references are clearly signaled one level deep, but `references/cancel-flow-playbook.md`, `references/dunning-guide.md`, and `scripts/churn_impact_calculator.py` are referenced yet absent — navigation leads nowhere, and supplementary detail that belongs in those files is inlined in SKILL.md. | 3 / 5 |
Total | 15 / 20 Passed |