Install
$ agentstack add skill-choshimwy-agentdevelopmentskills-ios-performance ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
iOS 性能分析与测试
Purpose
Analyze and improve iOS performance issues using the smallest credible evidence set, focusing on profiling strategy, baseline design, and before/after validation.
中文说明
该 Skill 负责性能基线设计、profiling 取证、模板选择、优化方向和前后对比验证。
- 适用于:掉帧、启动慢、CPU / 内存压力、性能回归、
measure(metrics:)、xctrace、Instruments。 - 不适用于:普通业务实现、泛化 crash 排查、普通单元/UI 测试补齐。
When to Use
- UIKit / SwiftUI 列表滚动掉帧、动画卡顿、页面进入慢、启动慢。
- CPU 高、内存增长、主线程繁忙、SwiftUI 更新过多。
- 需要建立
XCTest性能基线,如measure {}、measure(metrics:)、XCTApplicationLaunchMetric。 - 需要用
xcrun xctrace/ Instruments 录制Time Profiler、Animation Hitches、Allocations、Leaks、App Launch。
When Not to Use
- 问题核心是 crash、异常、野指针、对象未释放根因;使用
apple-debugging。 - 只是补业务单元测试、UI 测试或测试替身;使用
ios-feature-implementation(test-implementation)。 - 只是普通 SwiftUI/UIKit 功能实现;使用对应实现型 Skill。
Agent Rules
- Always define one symptom and one target interaction before profiling.
- Distinguish baseline design from runtime evidence collection.
- Prefer Release configuration and stable device / OS baselines for comparison.
- Output should include symptom, evidence, hypothesis, optimization direction, and validation method.
- When performance fixes edit code, document non-obvious performance invariants, caching/lifecycle side effects, concurrency boundaries, and fallback behavior in Chinese in touched code.
- Update stale comments, keep code comments Chinese by default, and avoid adding comments that merely restate optimization syntax.
- If the user asks for a formal HTML performance report, regression report, or shareable postmortem, prepare evidence summaries and route final document generation to
html-docs. - If code changes are produced, final closure still follows targeted validation / necessary verification plus independent reviewer subAgent
code-review; the implementation Agent must not self-review.
Inputs
{
"goal": "Analyze iOS performance issue",
"symptom": "scroll hitch | slow launch | cpu spike | memory growth | unknown",
"surface": "UIKit | SwiftUI | mixed",
"available_evidence": [],
"constraints": []
}
Outputs
{
"status": "completed | partial | blocked",
"symptom": "...",
"evidence": [],
"hypotheses": [],
"recommended_tools": [],
"optimization_directions": [],
"validation_plan": [],
"known_risks": [],
"next_action": "apple-verification | code-review | apple-debugging | ask-user | blocked"
}
Exit Conditions
completed: symptom, evidence path, optimization direction, and validation plan are all explicit.partial: useful hypotheses exist but evidence or environment is incomplete.blocked: no stable reproduction path, no usable profiling environment, or user intent is actually a different problem class.
Escalation Rules
- Escalate to
apple-debuggingwhen the issue is primarily crash, hang, leak root cause, or runtime behavior diagnosis. - Escalate to
ios-feature-implementation(test-implementation)when benchmark/regression test code is needed; escalate toapple-verificationfor low-cost targeted validation. - Escalate to
apple-docswhen official API or Instruments behavior must be confirmed. - Escalate to
html-docswhen the performance evidence and conclusions must become a formal HTML report or handoff document.
Token Budget
- Do not paste raw
xctracedumps or large logs. - Prefer local scripts or command-line aggregation to produce compact top stacks, time windows, thread hotspots, allocation deltas, and artifact paths before handing evidence to the Agent.
- For
.tracepackages, preferscripts/summarize-xctrace.pyto export selectedxctracetables and consumetrace-summary.json/trace-summary.mdinstead of raw XML exports. - Prefer compact symptom summaries, selected counters, and one clear evidence thread.
- Avoid mixing unrelated hotspots into the same report.
Reference Resources
scripts/summarize-xctrace.py: exports high-signalxctracetables and writes compacttrace-summary.json/trace-summary.mdevidence, including actionable findings, thread hotspots, stack signatures, and symbolication diagnostics.
Relationship to Other Skills
- Use
apple-debuggingfor crash and runtime fault analysis. - Use
ios-feature-implementation(test-implementation)for test authoring andapple-verificationfor targeted validation execution. - Use
ios-feature-implementationwithadvanced-swiftmode when performance work reveals deeper concurrency or abstraction implementation needs. - Use
apple-docsfor official API and Instruments fact lookup. - Use
html-docsfor final formal HTML performance reports; this Skill only owns the evidence, hypotheses, and validation plan.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: ChoshimWy
- Source: ChoshimWy/AgentDevelopmentSkills
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.