Conciseness

The body is a lean reference playbook: every line is a concrete command ('launchctl bootstrap gui/$(id -u) <plist>', 'xattr -d com.apple.quarantine <file>', 'codesign --force --deep --sign - <bundle>'), a technique note, or a defense pointer, with no tutorial padding explaining concepts like TCC or SIP that Claude already knows. Anti-drift check against anchor 4: there are no over-explained passages left to trim, so it fits the 'every token earns its place' anchor.

Actionability

Most techniques include copy-paste-ready commands with real flags (sqlite3 TCC.db queries, security dump-keychain, otool -l, osascript login item), matching the 'mostly executable guidance' anchor. It falls short of 5 because the LaunchAgent plist contents are never shown and dylib hijacking is described at the approach level without a complete executable chain; it is well above 3 since nothing is pseudocode.

Workflow Clarity

Sections are sequenced along a sensible attack lifecycle (Discovery → PrivEsc → Persistence → Evasion → Credential Access → C2 → Detection), but there are no validation checkpoints or feedback loops — e.g. no 'plutil -lint' the plist before bootstrapping, no 'launchctl print' to confirm the agent loaded, no verification that a quarantine attribute was actually removed. The destructive/batch guideline caps workflow clarity at 3 in the absence of validation steps.

Progressive Disclosure

The skill is a single well-sectioned ~75-line overview with no bundle files, so there is no nesting or buried content; header organization is clean and cross-references to sibling skills ('post-exploit/c2', 'reverser/ios-static') are clearly signaled. It is not a 5 because it exceeds a minimal overview — per-technique detail such as plist authoring and keychain cracking would sit naturally in one-level-deep reference files — but comfortably clears 3 since nothing that belongs in a separate file is inlined and navigation is easy.

Specificity

The description enumerates six concrete capabilities — 'launchd persistence, TCC bypass, Keychain credential access, dylib hijacking, Gatekeeper/quarantine evasion, and discovery' — which matches the anchor for multiple specific concrete actions with comprehensive coverage; a score of 4 would require noticeable coverage gaps, and none are evident for the macOS post-exploitation domain.

Completeness

The 'what' is explicit and concrete, but there is no 'Use when...' clause or equivalent explicit trigger guidance in the description; the judging guideline caps completeness at 3 in that case. It is not a 2 because the 'what' is clear and specific, and not a 4 because the 'when' is entirely absent rather than merely weakly implied.

Trigger Term Quality

Natural trigger terms are present ('macOS', 'launchd', 'TCC', 'Keychain', 'dylib hijacking', 'Gatekeeper', 'quarantine', 'Apple'), but common synonyms a user might say — 'mac', 'osx', 'macOS persistence' as a standalone phrase — are missing, matching the 'good keyword coverage; a few natural terms missing' anchor rather than the comprehensive synonym coverage of a 5.

Distinctiveness Conflict Risk

The description carves out a clear niche (macOS endpoint post-exploitation) with technology-specific triggers (launchd, TCC, Keychain, dylib, Gatekeeper) that are unlikely to fire for unrelated skills, matching the 'clear niche with distinct triggers; minimal conflict risk' anchor.