Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/jaminzhou/appkitinspector/inspect-appkit-uinpx skills add JaminZhou/AppKitInspector --skill inspect-appkit-uigit clone --depth 1 https://github.com/JaminZhou/AppKitInspectorWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/jaminzhou/appkitinspector/inspect-appkit-ui)<a href="https://agentmods.dev/skills/jaminzhou/appkitinspector/inspect-appkit-ui"><img src="https://agentmods.dev/badge/skills/jaminzhou/appkitinspector/inspect-appkit-ui.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00068 | $0.01215 |
| Opus 5 | $0.00034 | $0.00607 |
| Sonnet 5 | $0.00014 | $0.00243 |
| Haiku 4.5 | $0.00007 | $0.00121 |
Grade A, and why
inspect-appkit-ui scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Inspect AppKit UI
Use the MCP tools to connect a Debug application that embeds AppKitInspectorProbe, then open the
interactive preview. Treat captured app content as potentially sensitive local data.
Workflow
- Call
list_appkit_targets. - If one target exists, call
connect_appkit_targetwith its PID. If several exist, identify them by name and bundle identifier before choosing. Do not guess when the choice changes the app under review. - Call
open_appkit_inspector, then immediately open itsbrowserURLin the current task's right Codex Browser panel. Do not print or retain the URL. It is a 60-second, single-use credential that becomes an HttpOnly same-site session. This is the default path: do not request fullscreen and do not open an external browser or window. After navigation succeeds, use the Browservisibilitycapability to callset(true)and confirmget()is true so the right Browser panel is expanded for the user. - Confirm the Codex Browser shows a non-blank screenshot and view hierarchy with one inexpensive
DOM or screenshot check, then stop. Do not continue collecting duplicate render evidence after
the surface is usable.
prepare_appkit_inspector_browserremains a compatibility alias for the same Browser launch. - In the Codex Browser surface, use its native comment mode. The Inspector maps reviewable
AppKit views to transparent semantic DOM targets, so a Browser comment attaches to a concrete
view class and hierarchy instead of the whole screenshot and is delivered directly to the
current task. Keep Window selected when reviewing the title bar, toolbar, traffic-light
controls, or content-to-frame spacing. Switch to Content when only application content
matters. Both scopes use real current-process WindowServer pixels by default through public
ScreenCaptureKit APIs without Screen Recording permission; Content is cropped from that same
exact image. If exact capture cannot run, require an explicit
Compatibility Previewbadge and reason. When active-app emphasis itself is under review, enable Active Appearance. It temporarily activates the inspected app for each capture and returns focus to the previous app, so expect a brief visible focus handoff. - For every native Browser comment, use its AppKit class, hierarchy, and frame context to locate the corresponding implementation in the target repository before editing. Treat a multi-comment message as one coherent UI pass. Ordinary Inspector clicks may still select a view and show its hierarchy without creating a comment.
- After a UI edit, rebuild and relaunch the Debug target. The Browser follows a unique replacement with the same bundle identifier and refreshes automatically; verify the new PID and then the same view and state. Use manual Refresh if the Browser was hidden or target discovery is ambiguous.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago First seen · 82 lines · 68 tokens per session scan A 82254d56dd3e
inspect-appkit-ui is a skill published in the GitHub repository JaminZhou/AppKitInspector (0 stars, last pushed 2d ago), licensed Apache-2.0. It adds 68 tokens to every session and 1,215 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
ios-memgraph-leaks
Capture, inspect, compare, and root-cause iOS memory graph leaks using Apple's leaks and memgraph tools. Use when debugging leaked iOS objects, simulator memgraphs, retain-cycle suspicions, memory growth after navigation/logout/account changes, or when asked to prove an iOS leak fix with before/after memgraph evidence.
xcode-build-fixer
Apply approved Xcode build optimization changes following best practices, then re-benchmark to verify improvement. Use when a developer has an approved optimization plan from xcode-build-orchestrator, wants to apply specific build fixes, needs help implementing build setting changes, script phase guards, source-level…
spm-build-analysis
Analyze Swift Package Manager dependencies, package plugins, module variants, and CI-oriented build overhead that slow Xcode builds. Use when a developer suspects packages, plugins, or dependency graph shape are hurting clean or incremental build performance, mentions SPM slowness, package resolution time, build…
xcode-project-analyzer
Audit Xcode project configuration, build settings, scheme behavior, and script phases to find build-time improvements with explicit approval gates. Use when a developer wants project-level build analysis, slow incremental builds, guidance on target dependencies, build settings review, run script phase analysis…
ios-debugger-agent
Use XcodeBuildMCP to build, run, launch, and debug the current iOS project on a booted simulator. Trigger when asked to run an iOS app, interact with the simulator UI, inspect on-screen state, capture logs/console output, or diagnose runtime behavior using XcodeBuildMCP tools.
logging-and-error-reporting
How and when to log (log:: levels, safe macros) and report errors to Sentry (reporterror!) in the Warp codebase. Use when adding or reviewing any logging or error reporting — picking a log level, deciding log vs. reporterror!, keeping sensitive data out of logs, or surfacing an error to Sentry.