dpearson2699/swift-ios-skills / skills/ios-memgraph-analysis
ios-memgraph-analysis
VerifiedUse when capturing or analyzing an iOS .memgraph, especially when the task mentions a memory leak, heap growth, persistent memory increase, ownership path, or matched-capture comparison with Apple CLI tools. Covers unambiguous Simulator capture, leaks/heap/vmmap/malloc_history evidence, raw artifact preservation, and same-flow verification. Use debugging-instruments for interactive Xcode Memory Graph, Instruments, generic retain-cycle inspection, or LLDB work.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 9.4 KB |
| evals/evals.json | config | 4.1 KB |
| references/reachable-growth.md | markdown | 1.8 KB |
| scripts/capture_sim_memgraph.pyexecutable | script | 6.7 KB |
| scripts/summarize_memgraph.pyexecutable | script | 13.5 KB |
evals/evals.json
View on GitHub{
"skill_name": "ios-memgraph-analysis",
"evals": [
{
"id": 0,
"name": "dismissed-editor-ownership-proof",
"prompt": "EditorViewModel should deallocate after I dismiss an editor, but memory keeps growing. Design a Simulator memgraph investigation that proves the app-owned ownership path and verifies the eventual fix. I have two Simulators booted.",
"expected_output": "A lifetime-driven memgraph workflow that requires an explicit Simulator, captures one exact app process, preserves raw evidence, follows an app-owned type/path, and repeats the same flow after one ownership fix.",
"files": [],
"expectations": [
"Defines dismissal plus legitimate asynchronous cleanup as the expected EditorViewModel release boundary.",
"Requires an explicit Simulator UDID because two devices are booted and rejects ambiguous process matches.",
"Captures or exports a `.memgraph` after the release boundary and preserves the graph plus raw leaks output and manifest.",
"Starts from the app-owned EditorViewModel or its allocation stack and uses bounded trace/reference/grouped evidence to find the first app-controlled strong edge.",
"Distinguishes an unreachable cycle from a still-reachable abandoned owner.",
"Verifies by repeating the identical flow and querying the same type/path rather than relying on RSS or graph size."
]
},
{
"id": 1,
"name": "zero-leaks-persistent-growth",
"prompt": "`leaks` reports zero, but every open/close cycle adds about 8 MB and it never falls. What should I do with baseline and post-flow memgraphs? Include how vmmap, heap, traceTree/referenceTree, and Malloc Stack Logging fit together.",
"expected_output": "A correction-focused workflow that treats zero leaks as compatible with reachable growth, uses comparable pre/post graphs to locate and classify heap growth, and chooses ownership or allocation evidence based on Malloc Stack Logging availability.",
"files": [],
"expectations": [
"States that zero unreachable leaks does not rule out reachable but abandoned objects, caches, buffers, or heap regressions.",
"Uses matching baseline and post-flow graphs and `vmmap -summary` to locate the growing memory region.",
"Uses `heap --diffFrom=<baseline.memgraph> <post.memgraph>` to identify object types present in the post-flow heap but not the baseline.",
"Uses `heap --addresses=<pattern> <post.memgraph>` followed by `leaks --traceTree` or a grouped reference tree for ownership clues.",
"Uses a complete address-mode command such as `malloc_history post.memgraph -fullStacks <address>`, takes the address from `heap --addresses`, requires Malloc Stack Logging evidence, and keeps its setting identical across comparisons.",
"Preserves raw tool output and avoids treating text-parser columns as a stable API."
]
},
{
"id": 2,
"name": "weak-fix-claim-review",
"prompt": "I added `[weak self]` everywhere. The new graph file is smaller, RSS is down, and leaks reports 12 instead of 19 entries, so the retain-cycle bug is fixed. Review that conclusion and give me a proof standard.",
"expected_output": "A skeptical review that rejects aggregate metrics as ownership proof, requires a defined lifetime and app-owned path, evaluates weak/unowned semantics, and specifies same-flow repeated verification.",
"files": [],
"expectations": [
"Rejects graph file size, one RSS reading, and aggregate leak count as sufficient proof.",
"Requires naming the object, expected release event, and original app-owned ownership path or isolated cycle.",
"Warns that blanket `[weak self]` can hide lifecycle reasoning and may change required work semantics.",
"Uses weak only when the owner may legitimately disappear and unowned only with a proven lifetime guarantee.",
"Requires matching builds, settings, flows, cleanup wait, and repeated captures.",
"Calls the fix supported only when the same type/path or attributable pre/post growth disappears."
]
}
]
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |