ios-debugger-agent
Skill4 starsChanged 4 months agoMITView on GitHub
Build, launch, inspect, and drive iOS apps with the repository-configured XcodeBuildMCP server. Use on macOS for iOS Simulator builds, focused native test runs, semantic UI automation, screenshots, logs, or debugging, including T3 Code Mobile verification.
14 other projects use a version of this, 1 of them edited.
Adopted by
Most recently taken up by
- mohammadchehab/isnadibotedited, first seen Nov 15, 2026
- MMACABRE/SKILLStheir own version, first seen Nov 8, 2026
- spokvulcan/claude-code-build-ios-appstheir own version, first seen Oct 17, 2026
- openai/pluginstheir own version, first seen Oct 13, 2026
- Vvlladd/iOSskillsCollectiontheir own version, first seen Oct 13, 2026
- Samueladewole/claude-skillstheir own version, first seen Oct 13, 2026
Used with
Projects with it also use swift-concurrency-expert and app-store-changelog.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
eugenepyvovarovSkillswift-concurrency-expert
Swift Concurrency review and remediation for Swift 6.2+. Use when asked to review Swift Concurrency usage, improve concurrency compliance, or fix Swift concurrency compiler errors in a feature or file.
Also in 3 projects that use this
patrickserranoSkillapp-store-changelog
Create user-facing App Store release notes from git history. Use when asked to generate release changelog, App Store "What's New" text, or release notes based on git tags.
Also in 3 projects that use this
Other versions14 projects use a version of this.
Word for word · 2
- btsouth/toolport-studio★ 4100% same
- tsouth89/toolport-studio★ 4100% same
Edited · 1
- mohammadchehab/isnadibot★ 299% same
Wrote their own version · 11
- openai/plugins★ 7.8ksame name
- Dimillian/Skills★ 4ksame name
- omarshahine/HomeClaw★ 194same name
- robinebers/skills★ 48same name
- Adhamxon/claude-code-skills★ 2same name
- ParkerRex/Skills★ 2same name
- spokvulcan/claude-code-build-ios-apps★ 2same name
- MMACABRE/SKILLS★ 1same name
- tarunyadav1/ios-app-lifecycle★ 1same name
- Vvlladd/iOSskillsCollection★ 1same name
1 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository6 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .cursor/rules/cursor-cloud.mdcCursor rule
- ios-simulator-browserSkill
- test-t3-appSkill
- test-t3-mobileSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
21 rules. 0 are common, 5 are distinctive, and 16 are very specific to this project.
0
Common
Most of the 12 edited versions we read keep these.
5
Distinctive
This author's own choices. Worth a look.
16
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 8
Use the repository-configured xcodebuildmcp tools instead of requiring a globally installed Codex plugin.
- Distinctive: line 14
If the tools are missing:
- Very specific: line 18
Run npx --yes xcodebuildmcp@2.6.2 doctor when the server starts but simulator or UI-automation tools are unavailable.
- Very specific: line 21
Do not ask contributors to install the OpenAI build-ios-apps plugin globally.
- Very specific: line 25
Call session_show_defaults before discovery, build, launch, or UI work.
- Very specific: line 26
Call list_sims and select one explicit simulator UDID.
- Very specific: line 27
Call session_set_defaults with the project or workspace, scheme, Debug configuration, simulator ID, and bundle identifier when known.
- Distinctive: line 28
Keep every subsequent build, launch, screenshot, log capture, and UI action pinned to that same UDID.
- Distinctive: line 30
Avoid generic Mac window automation for switching among Simulator windows.
- Very specific: line 34
Use build_run_sim when native source, native dependencies, entitlements, or project configuration changed.
- Very specific: line 35
Use test_sim for the smallest relevant native test target or test cases.
- Very specific: line 36
Use launch_app_sim when a compatible app is already installed and no native rebuild is needed.
- Very specific: line 38
Do not run a build-only action immediately before build_run_sim unless the task requires both artifacts.
- Very specific: line 40
After launch, call snapshot_ui or screenshot before interacting.
- Very specific: line 44
Call snapshot_ui to obtain the current accessibility hierarchy and element references.
- Very specific: line 45
Use only current elementRef values whose snapshot entries list the intended action.
- Very specific: line 47
Use wait_for_ui for asynchronous transitions when available rather than fixed sleeps.
- Very specific: line 50
Use gesture or scoped swipe actions when needed.
- Very specific: line 54
Use start_sim_log_cap and stop_sim_log_cap with the exact bundle identifier for focused runtime logs.
- Distinctive: line 55
Use debugger tools only when the task requires runtime diagnosis.
- Distinctive: line 60
Stop only log captures, debugger sessions, apps, or simulators started for the current test.
“Common” is measured: kept by at least 6 in 10 of the 12 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
iOS Debugger Agent
Use the repository-configured xcodebuildmcp tools instead of requiring a globally installed Codex plugin. Prefer MCP tools over raw xcodebuild, xcrun, or simctl when the client exposes them.
Confirm availability
This workflow requires macOS 14.5 or newer, Xcode 16 or newer, and Node.js 18 or newer. The repository pins XcodeBuildMCP in both .mcp.json for Claude Code and .codex/config.toml for Codex. Project MCP servers may require one-time trust or approval followed by a new session.
If the tools are missing:
- Confirm the repository is trusted and its project MCP server was approved.
- Restart or recreate the agent session after approving configuration.
- Run
npx --yes xcodebuildmcp@2.6.2 doctorwhen the server starts but simulator or UI-automation tools are unavailable. Follow its actionable Xcode or AXe setup guidance. - Fall back to the pinned XcodeBuildMCP CLI or native Apple CLIs only when the current agent client cannot expose project MCP tools.
Do not ask contributors to install the OpenAI build-ios-apps plugin globally.
Establish one simulator context
- Call
session_show_defaultsbefore discovery, build, launch, or UI work. - Call
list_simsand select one explicit simulator UDID. Prefer a simulator that is already booted; boot an installed simulator when verification requires it, but do not create or download runtimes without user authorization. - Call
session_set_defaultswith the project or workspace, scheme, Debug configuration, simulator ID, and bundle identifier when known. - Keep every subsequent build, launch, screenshot, log capture, and UI action pinned to that same UDID.
Avoid generic Mac window automation for switching among Simulator windows. Explicit device identity is more reliable.
Choose build or launch
- Use
build_run_simwhen native source, native dependencies, entitlements, or project configuration changed. - Use
test_simfor the smallest relevant native test target or test cases; do not run an entire workspace test matrix routinely. - Use
launch_app_simwhen a compatible app is already installed and no native rebuild is needed. - To reuse an existing build artifact, use
get_sim_app_pathorget_app_bundle_id, install it withinstall_app_simwhen necessary, and then launch it. - Do not run a build-only action immediately before
build_run_simunless the task requires both artifacts.
After launch, call snapshot_ui or screenshot before interacting. An open Simulator window alone is not evidence that the intended app launched.
Drive the UI semantically
- Call
snapshot_uito obtain the current accessibility hierarchy and element references. - Use only current
elementRefvalues whose snapshot entries list the intended action. XcodeBuildMCP2.6.2does not accept coordinates fortap; when the app exposes no actionable reference, prefer a registered deep link or another app-supported route and otherwise report the accessibility blocker. - Refresh with
snapshot_uiafter navigation or layout changes. Element references are snapshot-specific. - Use
wait_for_uifor asynchronous transitions when available rather than fixed sleeps. - Capture a final
screenshotfor the state that proves the affected flow.
Use gesture or scoped swipe actions when needed. If a gesture is unreliable, return to a known route or relaunch rather than switching to generic desktop automation.
Capture logs and debug
- Use
start_sim_log_capandstop_sim_log_capwith the exact bundle identifier for focused runtime logs. - Use debugger tools only when the task requires runtime diagnosis; attach to the selected simulator and app rather than an ambiguous process.
- Summarize relevant errors instead of returning unbounded logs.
Clean up
Stop only log captures, debugger sessions, apps, or simulators started for the current test. Leave pre-existing simulators and unrelated sessions alone.
Upstream
Adapted from OpenAI's build-ios-apps plugin version 0.1.2 (ios-debugger-agent, MIT) and aligned with XcodeBuildMCP 2.6.2 tool names.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

