argent-native-profiler: a skill that profiles native CPU, UI hangs and memory leaks

argent-native-profiler makes your agent record a native performance trace on iOS or Android and report CPU hotspots, UI hangs and memory leaks.

Installs: 27.8K; on skills.sh
Stars: 3.1K; on GitHub
Updated: Oct 2026; last repo commit

The agent starts a native recording on a booted device, lets you or the device tools interact with the app, stops and exports the trace, then analyzes it. iOS uses Xcode Instruments through xctrace. Android uses Perfetto through adb shell perfetto. Results are sorted by severity: CPU hotspots, UI hangs and leaks.

It triggers for native-level performance problems. It needs the Argent profiler tools and a launched app on a booted device. On iOS it needs Xcode command-line tools on PATH. SKILL.md says simulator numbers reflect the host Mac, not real device hardware.

Good for

iOS and Android developers chasing UI hangs and jank
React Native developers who need native-side data next to React data
Developers hunting memory leaks

Install argent-native-profiler

Claude Code

Copy this prompt.

Install the argent-native-profiler skill in Claude Code. Run this command:

npx @swmansion/argent@latest init

Then check that it loaded and tell me how to use it.

2
Paste it into Claude Code. It shows the change and asks before running it.

Requires Node.js.

This runs the Argent installer. It sets up the Argent tools (an MCP server) that these skills call, plus the Argent skills, in the editors you pick. Skills copied on their own with the skills CLI have no tools to call. The installer offers to add Argent tools to your editor's auto-approve list. The default Automatic skills option installs all Argent skills, not only this one. Telemetry is on by default and argent telemetry disable turns it off.

From the project README. Source.

What's in the skill

When it loads

Native profiling for CPU hotspots, UI hangs, memory issues. iOS via xctrace; Android via Perfetto. Use when diagnosing native-level performance issues.

Sections in SKILL.md

1. Tools
2. Platform Support
3. Investigation Patterns
4. Workflow
Step 0: Ensure the target app is running
Step 1: Start recording
Step 2: Stop and export
Step 3: Analyze
Step 4: Present findings and ask about next steps
Step 5: Drill-down investigation
Step 6: Reload previous sessions
5. Understanding Results
6. Important Caveats

Key facts

Author: Software Mansion (Argent)
Repo: software-mansion/argent, path packages/skills/skills/argent-native-profiler
Installs on skills.sh: 27,849 (as of Nov 2, 2026)
GitHub stars: 3,138 (shared repo, Nov 2, 2026)
License: Apache-2.0 for source code; some bundled binaries are proprietary
Last repo commit: Oct 31, 2026
Files: 1 file: SKILL.md (no bundled scripts)
Needs: The Argent tools (MCP server), set up by the installer in the box above

What it includes

Starts recording on a booted device
iOS runs an xctrace recording for CPU, hangs and leaks. Auto-detects the running app and saves the trace to /tmp/argent-profiler-cwd/ with a timestamped name.
Runs code · native-profiler-start
Stops and exports trace data
iOS sends SIGINT to xctrace and exports XML. Android sends SIGTERM to the on-device perfetto daemon, then adb pulls the .pftrace file.
Runs code · native-profiler-stop
Returns a ranked bottleneck report
Groups findings as CPU hotspots, UI hangs and memory leaks, RED for the worst and YELLOW for the rest.
Read · native-profiler-analyze
Drills into hang stacks, callers, threads and leaks
Modes include hang_stacks, function_callers, thread_breakdown and leak_stacks.
profiler-stack-query
Lists and reloads earlier trace sessions
Used to compare before and after a fix.
Files · profiler-load
Optional leak attribution on iOS
Re-runs the start call so the app cold-launches with Malloc Stack Logging. It restarts the app and adds overhead.
Runs code · malloc_stack_logging
Seven steps from launching the app to reloading sessions
The agent must ask what to do after the report and must re-profile after any fix.
Workflow

From the official docs

Permissions

Low risk

It records traces from your running app and saves them to disk, using Xcode or adb tools. An optional leak mode restarts your app.

What it can do

Read your data
Create and change things
Delete things
Read-only mode

Watch out for

Installer can auto-approve Argent tools
`argent init` asks whether to add Argent tools to your editor's auto-approve list and its prompt defaults to yes. Say no if you want a confirmation on every tool call. github.com
Traces are saved in a temp folder
native-profiler-start saves the trace to /tmp/argent-profiler-cwd/. Android also pulls a .pftrace file to the host. Traces can contain function and library names from your app. github.com
Leak attribution restarts your app
malloc_stack_logging cold-launches the app with Malloc Stack Logging and adds overhead. SKILL.md says to use it only when you need leak attribution. github.com
Argent telemetry is on by default
The README says telemetry is opt-out. Run `argent telemetry disable` to turn it off; the privacy notice says it collects no source code, file paths or tool inputs, and records a hashed installation identifier derived from a device identifier, never the raw one. github.com

Technical details

Talks to	Argent telemetry (on by default)
Sign-in and keys	The skill text does not use an account or API key.

Questions

What does argent-native-profiler do?

It makes your agent record a native performance trace of your app on iOS or Android and report CPU hotspots, UI hangs and memory leaks, then drill into the causes.

How do I install it in Claude Code?

Use the install box above. It runs the Argent installer, which sets up the profiler tools this skill calls along with the skills.

What does it need on my machine?

On iOS it needs Xcode command-line tools on PATH; check with xcrun xctrace version. On Android it uses adb and Perfetto, and the target app must be debuggable, as SKILL.md notes.

Does it work on a real iPhone?

SKILL.md is not consistent: one line says a physical iPhone is not supported and to use a simulator, while the platform section mentions a connected device. Use a simulator to be safe.

Is it safe to run?

It records and saves traces from your running app and runs xctrace or adb tools. It bundles no scripts. The installer can also add Argent tools to your editor's auto-approve list, which you can decline.

Are the memory leak results reliable?

SKILL.md says most simulator leaks come back unattributed and are dominated by benign system allocations, so they are shown as a low-confidence summary. On Android the RSS-growth signal is a hint to confirm manually, not a confirmed leak.

How is it different from argent-react-native-profiler?

The React profiler covers React commits and JS CPU. This skill covers native CPU, hangs and leaks. SKILL.md says the React profiler skill starts native profiling alongside it by default.

Information

updated Nov 2, 2026

Repository: software-mansion/argent
Folder: packages/skills/skills/argent-native-profiler

software-mansion/argent skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 5

SKILL.md · github.com
Argent README · github.com
Argent installer: editor adapters · github.com
Argent installer: auto-approve step · github.com
init-skills.ts · github.com