Memorydetective MCP Server

memorydetective

Diagnose iOS retain cycles and performance regressions from your chat window. No Xcode required.

Highlights

• CLI-driven leak hunting. Read .memgraph files captured by Xcode (or by memorydetective itself on simulators), find ROOT CYCLEs, classify them against known SwiftUI/Combine patterns, and get a one-liner fix hint. All from a script or a chat.
• MCP-native. Plugs into Claude Code, Claude Desktop, Cursor, Cline, and any other MCP client. The agent drives the full investigate → classify → suggest-fix loop without you opening Instruments.
• Honest about its limits. No mocked outputs, no over-promises. Hangs analysis works clean from xctrace; sample-level Time Profile is parsed when xctrace symbolicates the trace and returns a structured workaround notice when it can't (the underlying xctrace SIGSEGV on heavy unsymbolicated traces is an Apple-side limitation we surface explicitly). Memory Graph capture works on Mac apps and iOS simulator; physical iOS devices still need Xcode.

What's new in v1.8 (2026-05-06): leaks --outputGraph regressed on macOS 26.x and aborts unless the target was launched with MallocStackLogging=1. This release fixes that end to end. captureMemgraph now emits a structured workaroundNotice when the regression hits, the new bootAndLaunchForLeakInvestigation tool absorbs build + boot + install + launch with the right env vars so capture works on the first try, and replayScenario + captureScenarioState close the verify-fix loop with deterministic before/after snapshots (memgraph + screenshot + UI tree, all labeled). 28 → 31 MCP tools, 213 → 287 tests.

Also recent: v1.7 (2026-05-03) added the 34th cycle pattern + fixTemplate snippets + compareTracesByPattern for trace-side CI gating. v1.6 (same day) added MCP Resources + Prompts. Full notes in CHANGELOG.

Quickstart

npm install -g memorydetective

Add to your MCP client config (Claude Code shown):

// ~/.claude/settings.json
{
  "mcpServers": {
    "memorydetective": { "command": "memorydetective" }
  }
}

Or use it directly from the command line:

memorydetective analyze   ~/Desktop/myapp.memgraph
memorydetective classify  ~/Desktop/myapp.memgraph

→ See Examples for chat-driven flows · API for the full tool reference · Configure for Claude Desktop / Cursor / Cline · USAGE.md for the full walkthrough including how fixes flow from diagnosis to your codebase.

---

Examples

Realistic chat snippets. The agent calls the MCP tools listed in API.

Diagnose a leak from a memgraph

You: I just exported a memory graph at ~/Desktop/example.memgraph. Something's leaking after I close a detail screen. What is it?

Claude: (calls analyzeMemgraph → classifyCycle) Found 4 ROOT CYCLE blocks across 60,436 leaks. The dominant one matches swiftui.tag-index-projection with high confidence. A closure attached via .tag() on a photo carousel ForEach is capturing self strongly through _viewModel.wrappedValue, which keeps DetailViewModel and the navigation coordinator alive after dismissal.

Suggested fix: replace [weak self] capture in the tap handler with a static helper, or weak-capture the view-model via [weak coord = self.coordinator]. Want me to find the exact file?

Verify a fix actually worked

You: I shipped the fix. Compare ~/Desktop/before.memgraph to ~/Desktop/after.memgraph and tell me whether DetailViewModel count dropped.

Claude: (calls diffMemgraphs → countAlive) DetailViewModel: 12 → 0 (delta -12). The swiftui.tag-index-projection cycle is gone. Total leak count fell from 60,436 to 55,576 (−4,860). Three other ROOT CYCLEs persist. They're SwiftUI internals, not your code.

Self-contained verify-fix loop (v1.8)

When leaks --outputGraph fails on macOS 26.x (Failed to get DYLD info for task), the cause is usually that the app was not launched with MallocStackLogging=1. The new orchestration tools handle this end to end without you wiring up xcodebuild + simctl + axe yourself.

You: Reproduce the leak in MyApp for me. Build, launch with the right env vars, repeat the photo carousel flow 5 times to amplify it, capture before, then I'll ship the fix and you capture after.

Claude: (calls bootAndLaunchForLeakInvestigation) Built + booted iPhone 15 (UDID ABCD-...). Launched MyApp (PID 49581) with MallocStackLogging=1. Bundle id com.example.MyApp.

(calls replayScenario with 5 taps on the carousel + back-button cycle, repeat 5) Executed 25 actions in 12.3s. No failures.

(calls captureScenarioState({label:"before"})) Wrote before.memgraph (63,201 leaks), before.png, before.ui.json. MallocStackLogging active, no workaroundNotice.

(you ship the fix and rebuild)

(calls bootAndLaunchForLeakInvestigation + replay + captureScenarioState({label:"after"}))

(calls diffMemgraphs → verifyFix) DetailViewModel: 15 → 0. swiftui.tag-index-projection PASS. Bytes freed: 28.4 MB.

Find the slow path on a real device

You: Profile DemoApp on my iPhone for 90 seconds and tell me where the hangs are.

Claude: (calls listTraceDevices → recordTimeProfile → analyzeHangs) Recorded ~/Desktop/run.trace (90s, attached to DemoApp on iPhone 17 Pro Max). Found 23 user-visible hangs (>500 ms each), longest 1.16 s, average 769 ms. Severe load. All on the main thread. Sample-level hotspots aren't symbolicated yet; open the trace in Instruments for that step (analyzeTimeProfile returns a structured workaround notice when xctrace can't symbolicate. See CHANGELOG known limits).

End-to-end: leak → file → fix suggestion

The full loop, from artifact to code edit. Combines memory-graph analysis with Swift source-bridging.

You: Tem um leak. Memgraph em ~/Desktop/example-leaks.memgraph. The cycle points at DetailViewModel. Find where it lives in the project ~/Development/myapp/ and suggest a fix.

Claude: (calls analyzeMemgraph → classifyCycle → reachableFromCycle) Found 4 ROOT CYCLE blocks. The dominant one matches swiftui.tag-index-projection with high confidence. reachableFromCycle confirms 4 DetailViewModel instances pinning ~1100 NSURLSessionConfigurations each. The VM is the culprit, the network stack is collateral.

(calls swiftSearchPattern to locate .tag( + onImageSliderTap patterns in ~/Development/myapp/Sources/) 1 match in MyApp/Views/MediaCarousel.swift:142.

(calls swiftGetSymbolDefinition({ symbolName: "DetailViewModel", projectRoot: "...", candidatePaths: ["MyApp/ViewModels/"] })) Declared at MyApp/ViewModels/DetailViewModel.swift:18. class DetailViewModel: ObservableObject.

(calls swiftFindSymbolReferences to scope the blast radius. 7 callsites)

Suggested fix at MediaCarousel.swift:142:

// before
.tag(index)
.onImageSliderTap { index in
    self.viewModel.handlePhotoTap(at: index)
}

// after. Static helper + weak captures
.tag(index)
.onImageSliderTap { [weak vm = _viewModel.wrappedValue,
                     weak coord = self.coordinator] index in
    Self.handlePhotoTap(index: index, viewModel: vm, coordinator: coord)
}

Want me to apply this and verify with a fresh .memgraph?

---

What it saves you

The pitch in one sentence: memorydetective turns a 50–500 MB binary memgraph (or a 200 KB leaks(1) text dump) into a 2–5 KB structured summary your AI agent can actually reason about. That changes the economics of using an LLM for iOS perf investigation.

Tokens (when paired with an AI agent like Claude / Cursor / Cline)

A real-world retain-cycle investigation, run twice. Once with memorydetective, once with the agent reading the raw leaks(1) output directly:

Translates to roughly $0.40–$1.20 per investigation depending on the model (Claude Opus / Sonnet / Haiku). Compounds linearly with file size and investigation depth.

Developer time

The same investigation, measured by the developer:

Numbers are rounded from a single anonymized real investigation (a SwiftUI retain cycle over a tagged ForEach that pinned ~28 MB of network-stack state). Your mileage will vary with cycle complexity and codebase size.

When the win is marginal

Be honest about where this doesn't help much:

• Tiny memgraphs (a single cycle, < 50 KB raw): MCP overhead is roughly token-neutral vs. Raw read. The dev-time win still holds (no manual cycle parsing) but the token win shrinks.
• One-shot symbol lookups without a leak attached: just use grep, you don't need this.
• First-time investigations on a new codebase: the agent still needs orientation turns regardless of MCP. The compounding wins kick in on the second and later investigations once the agent has cached the project's shape.

The win compounds with (a) file size, (b) investigation depth (multi-turn), and (c) how many leaks you investigate per quarter. For a single dev fixing one leak per year, the value is mostly the dev-time saving. For a team running CI gates with verifyFix across every PR, the token + time savings stack across hundreds of runs.

---

Configure

The memorydetective binary speaks MCP over stdio. Point any MCP-compatible client at it.

// ~/.claude/settings.json (global) or .mcp.json (per-project)
{
  "mcpServers": {
    "memorydetective": { "command": "memorydetective" }
  }
}

// ~/Library/Application Support/Claude/claude_desktop_config.json
{
  "mcpServers": {
    "memorydetective": { "command": "memorydetective" }
  }
}

Restart Claude Desktop after editing.

// ~/.cursor/mcp.json
{
  "mcpServers": {
    "memorydetective": { "command": "memorydetective" }
  }
}

// VS Code settings.json
{
  "cline.mcpServers": {
    "memorydetective": { "command": "memorydetective" }
  }
}

Kiro supports MCP servers via its global config. The block mirrors Claude Desktop's:

{
  "mcpServers": {
    "memorydetective": { "command": "memorydetective" }
  }
}

Consult Kiro's MCP setup docs for the exact config file path on your system.

GitHub Copilot supports MCP servers in Agent mode (VS Code 1.94+). Add to .vscode/mcp.json in your repo:

{
  "servers": {
    "memorydetective": {
      "type": "stdio",
      "command": "memorydetective"
    }
  }
}

Copilot's MCP integration moves fast. If this snippet is stale, see the VS Code MCP docs.

---

API

31 MCP tools + 34 Resources + 5 Prompts, grouped by purpose. Tool descriptions are tagged with a category prefix ([mg.memory], [mg.trace], [mg.build], [mg.scenario], [mg.code], [mg.log], [mg.render], [mg.ci], [mg.discover], [meta]) so related tools are visible at a glance.

Many tools include a suggestedNextCalls field in their response. A typed list of { tool, args, why } entries pre-populated from the current result, so the orchestrating LLM can chain calls without re-reasoning. Start with getInvestigationPlaybook(kind) for the canonical sequence. Or just type /investigate-leak (one of the Prompts) in any client that exposes MCP slash commands.

The cycle classifier ships 34 named antipatterns spanning SwiftUI (including the Swift 6 / @Observable / SwiftData / NavigationStack era), Combine, Swift Concurrency (incl. AsyncSequence-on-self and the new Observations API), UIKit (Timer/CADisplayLink/UIGestureRecognizer/KVO/URLSession/WebKit/DispatchSource), Core Animation, Core Data, Coordinator pattern, and the popular third-party libs RxSwift + Realm. Each pattern carries:

• a textual one-line fixHint
• a confidence tier (high / medium / low)
• a staticAnalysisHint pointing at the SwiftLint rule that complements the runtime evidence (or an explicit gap notice when no rule exists. Reinforces the differentiator: memorydetective sees what linters miss at parse time)
• a fixTemplate with concrete Swift before/after snippets (new in v1.7) the agent can adapt directly to the user's code via the SourceKit-LSP source-bridging tools

Read & analyze (13)

Capture / record (3)

Verify-fix orchestration (3, v1.8)

These three tools combine into a single deterministic verify-fix loop: launch the app with MallocStackLogging=1 so leaks works, drive the UI to amplify the suspected leak, snapshot before, ship the fix, snapshot after, then diffMemgraphs.

Discover (2)

Render (1)

CI / test integration (2)

Swift source bridging (5)

Pair the memory-graph diagnosis with source-code lookups via SourceKit-LSP. Closes the loop "found this leak in the cycle → find the file/line in your project".

These tools require macOS + Xcode (full Xcode, not just Command Line Tools. xcrun sourcekit-lsp must be available). They start a sourcekit-lsp subprocess per project root and reuse it across calls; the subprocess shuts down after a 5-minute idle window.

Why captureMemgraph doesn't work on physical iOS devices: leaks(1) only attaches to processes running on the local Mac (which includes iOS simulators). Memory Graph capture from a real device goes through Xcode's debugger over USB/lockdownd. Different mechanism, no public CLI equivalent.

Resources (34)

The cycle-pattern catalog is also surfaced as MCP resources, browsable at memorydetective://patterns/{patternId}. Each resource is a markdown body with the pattern name, a longer description, and the fix hint. Use this to let an agent (or a human in a UI-aware MCP client) browse the catalog without burning a classifyCycle call.

memorydetective://patterns/swiftui.tag-index-projection
memorydetective://patterns/concurrency.async-sequence-on-self
memorydetective://patterns/webkit.wkscriptmessagehandler-bridge
memorydetective://patterns/swiftdata.modelcontext-actor-cycle
…

resources/list returns all 34 entries. resources/read resolves any memorydetective://patterns/{id} URI to its markdown body.

Prompts (5)

Investigation playbooks are exposed as MCP prompts (slash commands in clients that surface them, e.g. Claude Code).

Each prompt fills the canonical playbook's argument templates with the user-provided values, then hands the agent a ready-to-execute brief. Calls the same tools listed in Read & analyze. Prompts are an orchestration shortcut, not a separate engine.

CLI mode

The same binary is also a thin CLI for scripting and CI:

memorydetective analyze   <path-to-.memgraph>    # totals, ROOT CYCLEs, diagnosis
memorydetective classify  <path-to-.memgraph>    # match patterns + render fix hint
memorydetective --help
memorydetective --version

When called with no arguments, the binary starts as an MCP server over stdio.

---

Requirements

• macOS with Xcode Command Line Tools (xcode-select --install)
• Node.js ≥ 20

Develop

git clone https://github.com/carloshpdoc/memorydetective
cd memorydetective
npm install
npm test                  # 61 unit tests
npm run build             # build → dist/
npm run dev               # tsx, stdio mode (dev mode)
./scripts/demo.sh         # full demo against a real .memgraph (set MEMGRAPH=path)

Contributing

Contributions are welcome. Bug reports, feature requests, new cycle patterns, all of it.

• Bugs / feature requests: open an issue.
• PRs: fork → branch → npm install → make changes → npm test (206 tests must stay green) → open a PR with a concise description of what changed and why.

Adding a cycle pattern to classifyCycle

classifyCycle ships with 34 built-in patterns covering SwiftUI (incl. Swift 6 / @Observable / SwiftData / NavigationStack), Combine, Swift Concurrency (incl. AsyncSequence-on-self and Observations), UIKit (Timer / CADisplayLink / UIGestureRecognizer / KVO / URLSession / WebKit / DispatchSource), Core Animation, Core Data, the Coordinator pattern, RxSwift, and Realm. To add one:

1. Edit src/tools/classifyCycle.ts. Add an entry to PATTERNS with id, name, fixHint, and a match function.
2. Add a test in src/tools/readTools.test.ts that asserts the new pattern fires against a representative memgraph fixture.
3. Add a staticAnalysisHint entry in src/runtime/staticAnalysisHints.ts (the test in that file enforces 1:1 coverage with PATTERNS).
4. Add a fixTemplate entry in src/runtime/fixTemplates.ts (same 1:1 coverage guard).
5. Open a PR.

Support this project

If memorydetective saves you time, you can support continued development:

• ☕ Buy me a coffee
• 💖 Sponsor on GitHub

Every contribution helps keep this maintained and documented.

License

Apache 2.0. See LICENSE and NOTICE.

Permits commercial use, modification, distribution, patent use. Includes attribution clause via the NOTICE file.

Why "memorydetective"?

Hunting retain cycles in SwiftUI feels like detective work: you have a body (the leaked instance), a crime scene (the .memgraph), and a chain of suspects (the retain chain). The tool helps you read the evidence and name the killer. The brand follows the work.