Content
52%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 delivers a clear, well-sequenced modeling workflow with genuinely useful startup-specific benchmarks and validation checks, but it is bloated with standard finance knowledge Claude already has, and its progressive-disclosure promise is broken: every referenced reference/example file is missing from the bundle while their would-be content is inlined. Tightening the overview and actually shipping the referenced files would fix the two weakest dimensions.
Suggestions
Actually create the referenced files (`references/model-templates.md`, `references/unit-economics.md`, `references/fundraising-scenarios.md`, and the three `examples/` files) and move the inlined Business Model Templates and step-level detail into them, or remove the dead references; also reconcile the inconsistent `resources/implementation-playbook.md` path to `references/`.
Trim sections explaining concepts Claude already knows (MRR/ARR definitions, basic dilution math, COGS/S&M/R&D/G&A category definitions) and the boilerplate "Use this skill when / Do not use this skill when" sections, keeping only the startup-specific heuristics (typical ratios, retention curves, pitfalls, sanity-check thresholds).
Convert the Model Validation sanity checklist into a feedback loop — e.g., "If LTV/CAC < 3 or payback > 18 months, revisit churn and CAC assumptions in Steps 2-3, then re-run the checklist" — so failed checks drive concrete next steps.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~465-line body spends large sections on concepts Claude already knows ("ARR = MRR × 12", "Post-Money = Pre-Money + Investment", definitions of COGS/S&M/R&D/G&A, CAC/LTV) plus boilerplate padding ("Use this skill when"/"Do not use this skill when" sections, "Apply relevant best practices and validate outcomes"). It avoids the severely-padded tutorial style of the 1 anchor only because the domain-specific heuristics (retention curves, pitfalls, sanity-check thresholds) carry real value. | 2 / 5 |
Actionability | Mostly concrete, executable guidance: worked formulas ("Runway = Ending Cash / Average Monthly Burn"), worked numbers ("Engineer: $150K salary × 1.35 = $202K fully-loaded"; "$5M at $20M pre-money → Dilution: 20%"), specific benchmarks ("LTV/CAC > 3, payback < 18 months"), and a 9-step sequence. Not a 5 because some sections are just consideration lists ("Clarify revenue model and pricing") and the complete worked examples are deferred to files that don't exist in the bundle. | 4 / 5 |
Workflow Clarity | A clearly sequenced 7-step process, a 9-step Quick Start, and an explicit "Model Validation" sanity checklist ("[ ] Revenue growth rate is achievable (3x in Year 2, 2x in Year 3)") give most needed checkpoints. Not a 5 because there is no feedback loop — nothing instructs what to revise when a sanity check fails. | 4 / 5 |
Progressive Disclosure | The body signals references to `references/model-templates.md`, `references/unit-economics.md`, `references/fundraising-scenarios.md`, and three `examples/` files — none of which exist in the bundle (which contains only SKILL.md) — and also points to an inconsistent `resources/implementation-playbook.md` path. Meanwhile the template-level detail that belongs in those files (four business-model templates, deep step-by-step material) is fully inlined, matching the anchor "content that clearly belongs in separate files is inlined". | 2 / 5 |
Total | 12 / 20 Passed |