Content
82%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 strong, execution-ready cheat sheet: nearly every line is a runnable command with a domain-specific pointer of what to look for, organized in a clear end-to-end workflow. Minor room to improve by trimming tool-selection commentary and adding recovery guidance for common failure cases.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Nearly lean throughout: commands annotated with domain-specific facts Claude may not know (cryptid flags, ATS/plist keys, keychain attributes). A few lines of tool commentary ('Hopper Disassembler — UI-focused, decent Obj-C support', 'IDA Pro — best Objective-C / Swift but $$$') could be trimmed. | 4 / 5 |
Actionability | Copy-paste-ready bash commands cover the full workflow: acquisition ('ipatool download -b com.target.app', 'frida-ios-dump'), unpacking, triage ('security cms -D -i ... | plutil -p -', 'otool -l ... | grep -A4 LC_ENCRYPTION_INFO'), class-dump, disassembly (r2 command sequence), and concrete grep patterns for secrets, endpoints, and Firebase configs. | 5 / 5 |
Workflow Clarity | Clear logical sequence (Acquire → Unpack → numbered Quick triage → Class-dump → Disassembly → Things to look for) with a real checkpoint (cryptid 0 = ready to analyze). Read-only analysis, so the destructive/batch validation cap does not apply, but there are no feedback loops for failure cases such as class-dump producing nothing on a Swift-only binary. | 4 / 5 |
Progressive Disclosure | Single well-sectioned file with no bundle files to reference; external reading is cleanly collected in a References section and the cross-skill pointer to mobile/SKILL.md is one level deep. At ~150 lines, some 'Things to look for' subsections (secrets, JS bridges, file storage) could be split into reference files, but nothing is buried. | 4 / 5 |
Total | 17 / 20 Passed |