菜单
Profile native macOS/iOS apps for CPU hotspots, hangs, and hitches using Instruments traces. Use when asked to identify performance hotspots, profile CPU usage, or diagnose slow code paths without opening Instruments.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Consult a stronger model mid-task for a second opinion before committing to an approach, when stuck, or before declaring a non-trivial task done — the fast-executor + strong-advisor pattern, implemented with the Agent and Workflow tools (Claude Code does not expose Anthropic's raw API advisor tool). Use when running on a cheaper/faster model for mechanical work and a hard design/architecture/risk decision needs a stronger check, or when authoring a Workflow script that should get expert review partway through instead of only at the end.
Use when the user says "caveman mode", "talk like caveman", "use caveman", "less tokens", or "be brief", invokes /caveman, or when token efficiency is requested. Switches responses to an ultra-compressed terse style (multiple intensity levels available, ~75% token reduction) while preserving full technical accuracy.
Use right after being corrected on something a CLAUDE.md file should already have told you, when explicitly asked to review CLAUDE.md, or periodically after a session with a lot of back-and-forth — reviews THIS conversation for friction and proposes a bounded CLAUDE.md edit. A lightweight, single-session complement to skill-tuning-loop (which mines many sessions and validates a proposal against held-out cases before surfacing it — reach for that instead when the ask is "make this systematic," not "check right now"). Own version of the dx plugin's `/dx:review-claudemd`.
Generate a solution, evaluate it against explicit pass/fail criteria, refine using the feedback, repeat until it passes or a round cap is hit. Use for tasks with a clear, checkable bar and demonstrable value from iteration — a bug fix that must pass a test suite, code that must satisfy a lint/style rubric, a document that must meet stated requirements. Distinct from advisor-checkpoint (one strategic consult before committing) — this is a convergence loop against a concrete standard.
Use when PR checks fail, CI is red, or GitHub Actions workflows break - systematically inspects failing checks via gh CLI, pulls logs, checks for flakiness and scopes the breaking commit, scopes external checks, then creates fix plan using existing plan skill
Write a structured handoff document before a session ends, before a context-heavy interruption (SSH/tmux disconnect, context running low, a worktree being handed to another worker or agent), or when explicitly asked for a handoff — captures progress, what worked/failed, and the exact next step so a fresh session can resume without re-deriving context. Own version of the dx plugin's `/dx:handoff`.
| name | native-app-profiling |
| description | Profile native macOS/iOS apps for CPU hotspots, hangs, and hitches using Instruments traces. Use when asked to identify performance hotspots, profile CPU usage, or diagnose slow code paths without opening Instruments. |
This is a redirect skill. This lacquer's Instruments trace recording and analysis workflow lives in swiftui-expert-skill — see its "Record a new Instruments trace" and "Trace-driven improvement" task-workflow sections, backed by scripts/record_trace.py and scripts/analyze_trace.py and documented in full in references/trace-recording.md and references/trace-analysis.md. That workflow already covers everything this skill used to teach manually: template selection by device kind, attach/launch/stop-file recording modes, and analysis with automatic symbolication, window scoping, main-thread coverage (main_running_coverage_pct), and SwiftUI cause-graph fan-in (--fanin-for).
Do not use raw xcrun xctrace, atos, vmmap, or pgrep for this. FlowDeck's ban on raw Apple CLI tools (xcodebuild, xcrun, simctl, devicectl) extends to Instruments tooling in this lacquer for the same reason — record_trace.py/analyze_trace.py wrap xctrace and handle symbolication automatically, so the manual atos/vmmap load-address dance this skill used to document is no longer part of the workflow.
Any request to profile CPU usage, find performance hotspots, or diagnose hangs/hitches in a native iOS/macOS app — whether or not the code under investigation is SwiftUI. swiftui-expert-skill's trace workflow is not SwiftUI-specific for recording or the Time Profiler/Hangs/Animation Hitches lanes: only the swiftui and swiftui-causes analysis lanes are SwiftUI-specific, and they simply report available: false on non-SwiftUI code paths.
These are the two items from the old manual workflow that remain relevant and aren't already documented in swiftui-expert-skill's trace references:
xctrace (wrapped by record_trace.py) may prompt for Developer Tools access or Full Disk Access on first use on a given machine, or require elevated privileges for certain system-wide captures. This is separate from the per-device signing/trust failure mode already documented in references/trace-recording.md.Everything else — device/template selection, starting or stopping a recording, scoping analysis to a time window, interpreting coverage percentages, and mapping findings back to source files — is swiftui-expert-skill's job. Follow its workflow directly rather than duplicating it here.
Every capability this skill used to teach (record, export, symbolicate, analyze) now has a strictly better equivalent in swiftui-expert-skill. This skill was kept as a short redirect rather than deleted outright, since removing a skill entirely is a bigger call than trimming its content — flagged here for a human decision on whether to fully retire native-app-profiling in a future pass.