New: 98,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

Loupe

by heoblitz in heoblitz/Loupe

Inspect, act on, and verify native Apple-platform app interfaces using Loupe runtime snapshots, accessibility, screenshots, and action traces.

MITAuto-check passedMobile

Install Loupe

skills CLI
$ npx skills add heoblitz/Loupe --skill loupe -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install heoblitz/Loupe loupe --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/heoblitz/Loupe.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/loupe .claude/skills/loupe && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
loupe
GitHub stars
104
Token cost
~709 tokens
SKILL.md length
315 words
Files
4 (incl. references)
Skills in repo
1
Repo updated
First seen
Licence
MIT

At a glance

Inspect, act on, and verify native Apple-platform app interfaces using Loupe runtime snapshots, accessibility, screenshots, and action traces.

  • Runtime UI diagnosis
  • SKILL.md covers Choose The Runtime, Short Working Loop, Keep Context Small and Task-Specific References
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
  • Design comparison

What it does

Loupe is an agent skill from heoblitz/Loupe. Inspect, act on, and verify native Apple-platform app interfaces using Loupe runtime snapshots, accessibility, screenshots, and action traces. Use for runtime UI diagnosis, design comparison, simulator injection, or linked LoupeInjector workflows.

Its SKILL.md is about 710 tokens, which your agent loads only when the skill is triggered. The skill folder holds 4 other files, including reference files (for example `references/actions-and-mutations.md`, `references/evidence-workflow.md` and `references/runtime-modes.md`).

It sits in Mobile, covering iOS development and Accessibility. The repository describes itself as: Swift CLI that helps AI coding agents inspect runtime UI in Apple platform apps. The licence is MIT.

When your agent uses it

  • Runtime UI diagnosis
  • Design comparison
  • Simulator injection
  • Linked LoupeInjector workflows

Example prompts

  • “/loupe”

What it can do on your machine

Read from SKILL.md and the folder at commit f831291. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md (its code samples are bash).

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Loupe loads about 709 tokens when it runs, and up to ~4.9k if it reads all its reference files. Until then it costs about 63 tokens; SKILL.md has 315 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~63
When it runs · the whole SKILL.md, loaded when a task matches
~709
With references · SKILL.md plus every file in references/, read only if the agent opens them
~4.9k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from heoblitz/Loupe at commit f831291, republished under its MIT licence (© heoblitz). 315 words, ~709 tokens.

Download SKILL.mdSave it as .claude/skills/loupe/SKILL.md (or your agent's skills folder). This skill also uses 3 other files; get the full folder from GitHub.
name
loupe
description
Inspect, act on, and verify native Apple-platform app interfaces using Loupe runtime snapshots, accessibility, screenshots, and action traces. Use for runtime UI diagnosis, design comparison, simulator injection, or linked LoupeInjector workflows.

Loupe

Choose The Runtime

  • Identify the platform and attachment mode. Prefer the host printed by loupe app launch; app current may refer to another app.
  • Keep the same --host <url> or --bundle-id <id> across live commands. Add --udid <id> when the same bundle runs on multiple simulators.
  • Simulator injection needs no app source change or import LoupeKit. Physical devices need a debug-only linked and embedded LoupeInjector.
  • Read runtime-modes for launch and platform setup.

Short Working Loop

bash
loupe act targets --host <host> --search <label-or-id>
loupe act tap '#1' --host <host>
loupe ui query --host <host> --tree accessibility --test-id <expected-id>
  • Aliases are quoted, one-shot intents. List targets again after an alias action. If the app-owned testID is known, use act tap --test-id <id> directly.
  • Discover through accessibility; inspect layout/style through the view tree. Zero matches means not found; multiple matches require a more precise target.
  • Refs belong to a snapshot/session. Use --snapshot <file> for saved-ref actions and mutations. Text is for discovery, not tap targeting.
  • Verify the expected change with a focused query, wait, value, screenshot, or trace. Successful dispatch alone does not prove the intended result. If dispatch may already have occurred, inspect state before retrying.

Keep Context Small

  • For layout/design work, capture ui report --host <host> --output <dir>. Keep the snapshot and images on disk, then use ui query <snapshot.json> or ui node <snapshot.json> --ref <ref> --fields node.
  • ui tree defaults to 80 lines with bounded line lengths. Focus with --ref or --limit; use --all only when full output is needed as an artifact.
  • Use ui tree --text, ui screen, ui paint, debug object-graph, and debug defaults (also for feature flags). Check subcommand help only for unfamiliar options; avoid repeatedly loading help or full JSON/build logs.
  • For design iteration, review one report/screenshot, compare with the design, try a small mutation if useful, then patch source and verify after relaunch.

Task-Specific References

Read only the reference relevant to the next operation:

  • actions-and-mutations: input, waits, traces, mutation, self-sizing, and source reflection.
  • evidence-workflow: visibility/occlusion, design comparison, SwiftUI/probes, and focused diagnostics.

© heoblitz, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 3 other files (references) in skills/loupe of heoblitz/Loupe.

  • SKILL.md
  • references/actions-and-mutations.md
  • references/evidence-workflow.md
  • references/runtime-modes.md

Open the folder on GitHubat commit f831291

Compare with similar skills

Loupe next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Loupe compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Loupe this skillheoblitz/Loupe104—~709Automated safety check: PassMIT
Hig Components Contentraintree-technology/hig-doctor1435 repos~1.3kAutomated safety check: PassMIT
iOS Accessibilitydpearson2699/swift-ios-skills1.2k—~4.6kAutomated safety check: PassCustom licence
Assistive Accessgustavscirulis/snapgrid1171 repos~2.5kAutomated safety check: PassCustom licence
Generatorsgustavscirulis/snapgrid1171 repos~4kAutomated safety check: NotesCustom licence
UI Review TahoeKartikLabhshetwar/better-shot2.4k2 repos~1.7kAutomated safety check: PassCustom licence

Similar skills

  • Hig Components Content

    raintree-technology/hig-doctor

    Apple Human Interface Guidelines for content display components.

    143 GitHub starsUsed in 5 repos~1.3k tokens
    MobileAuto-check passed

  • iOS Accessibility

    dpearson2699/swift-ios-skills

    Build and audit SwiftUI, UIKit, and AppKit accessibility for VoiceOver, Voice Control, Switch Control, Full Keyboard Access, Dynamic Type, focus restoration, labels/traits/actions, traversal, custom…

    1.2k GitHub stars~4.6k tokensUpdated 2 mo ago
    MobileAuto-check passed

  • Assistive Access

    gustavscirulis/snapgrid

    Assistive Access implementation for cognitive accessibility including simplified scenes, navigation icons, runtime detection, and design principles.

    117 GitHub starsUsed in 1 repo~2.5k tokens
    MobileAuto-check passed

  • Generators

    gustavscirulis/snapgrid

    Code generator skills that produce production-ready Swift code for common app components.

    117 GitHub starsUsed in 1 repo~4k tokens
    MobileAuto-check: notes

  • UI Review Tahoe

    KartikLabhshetwar/better-shot

    Comprehensive UI/UX review for macOS Tahoe apps. An agent skill from KartikLabhshetwar/better-shot.

    2.4k GitHub starsUsed in 2 repos~1.7k tokens
    Frontend & DesignAuto-check passed

  • iOS Simulator Skill

    n0an/VivaDicta

    21 production-ready scripts for iOS app testing, building, and automation.

    131 GitHub starsUsed in 1 repo~2.5k tokens
    MobileAuto-check passed

Categories

Questions about Loupe

What does Loupe do?

Inspect, act on, and verify native Apple-platform app interfaces using Loupe runtime snapshots, accessibility, screenshots, and action traces. Loupe is an agent skill from heoblitz/Loupe. Inspect, act on, and verify native Apple-platform app interfaces using Loupe runtime snapshots, accessibility, screenshots, and action traces.

When should I use Loupe?

Loupe fits situations like: runtime UI diagnosis; design comparison; simulator injection; linked LoupeInjector workflows.

How do I install Loupe in Claude Code?

Run `npx skills add heoblitz/Loupe --skill loupe -a claude-code`. Or copy the skill folder (skills/loupe in heoblitz/Loupe) into .claude/skills/loupe in your project. Claude Code loads it when a task matches its description.

How do I install Loupe in Codex?

Run `npx skills add heoblitz/Loupe --skill loupe -a codex`. Or copy the skill folder (skills/loupe in heoblitz/Loupe) into .agents/skills/loupe in your project. Codex loads it when a task matches its description.

Can I use Loupe in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add heoblitz/Loupe --skill loupe -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/loupe, .gemini/skills/loupe, .github/skills/loupe and .opencode/skills/loupe in your project.

What does Loupe need to run?

SKILL.md names no scripts, command-line tools or credentials: Loupe is instructions for the agent only.

Does Loupe access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Loupe safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Loupe use?

Loupe is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Loupe use?

About 709 tokens (SKILL.md is roughly 2.8k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 4.2k tokens, read only when the agent opens those files.

What are the alternatives to Loupe?

Skills that share tags, products or a category with Loupe: Hig Components Content (raintree-technology/hig-doctor, 143 stars), iOS Accessibility (dpearson2699/swift-ios-skills, 1.2k stars), Assistive Access (gustavscirulis/snapgrid, 117 stars) and Generators (gustavscirulis/snapgrid, 117 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Loupe?

heoblitz (a GitHub user) maintains it in heoblitz/Loupe, which has 104 GitHub stars. The repository was last updated on October 2, 2026.

Source: heoblitz/Loupe on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.