Content
90%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 dense, actionable audit skill that assumes Claude's expertise and provides concrete commands, API identifiers, and a structured output template. The main gaps are the absence of an explicit numbered workflow with validation feedback loops and the opportunity to split the category checklists into reference files.
Suggestions
Add a short numbered workflow at the top (e.g., 1. confirm authorization, 2. select MASVS profile, 3. run static analysis, 4. run runtime tests, 5. verify each finding, 6. produce output) with explicit validation checkpoints.
Move the per-category MASVS checklists into separate reference files (e.g., references/masvs-storage.md) referenced one level deep from SKILL.md to improve progressive disclosure and reduce the inline density.
Add a brief verification/feedback loop note per checklist category so findings are re-checked after remediation rather than reported once.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean checklist-style body that assumes Claude's competence — it lists concrete checks and commands without explaining what MASVS, crypto, or IPC are, and every section earns its tokens. | 5 / 5 |
Actionability | Highly actionable with specific commands ("strings <app>.ipa", "otool -hv", "readelf -h", "adb backup"), concrete API identifiers (kSecAttrAccessible, EncryptedSharedPreferences, BiometricPrompt), and a tool table covering common cases. | 5 / 5 |
Workflow Clarity | Clear section sequence (authorization gate → MASVS checklists → static tools → runtime testing → output format) with an authorization checkpoint and per-finding verification field, but lacks an explicit numbered end-to-end sequence with validation feedback loops. | 4 / 5 |
Progressive Disclosure | Well-organized single-file overview with clearly signaled sections and external-standard references, but the per-category checklists are dense and could be split into one-level-deep reference files for easier navigation. | 4 / 5 |
Total | 18 / 20 Passed |