Agent skill

Idevice

by exelearning in exelearning/exelearning

Create or modify an eXeLearning iDevice, including edition/export data, legacy compatibility, assets, TypeScript bundles and teardown.

AGPL-3.0Auto-check passed

Install Idevice

skills CLI
$ npx skills add exelearning/exelearning --skill idevice -a claude-code

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

GitHub CLI
$ gh skill install exelearning/exelearning idevice --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/exelearning/exelearning.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/idevice .claude/skills/idevice && 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
idevice
GitHub stars
145
Token cost
~1.7k tokens
SKILL.md length
677 words
Files
1
Skills in repo
17
Repo updated
First seen
Licence
AGPL-3.0

At a glance

Create or modify an eXeLearning iDevice, including edition/export data, legacy compatibility, assets, TypeScript bundles and teardown.

  • SKILL.md covers Trace the complete lifecycle, Edition lifecycle: own what… and TypeScript/bundled devices
  • Calls bun

What it does

Idevice is an agent skill from exelearning/exelearning. Create or modify an eXeLearning iDevice, including edition/export data, legacy compatibility, assets, TypeScript bundles and teardown.

Its SKILL.md is about 1.7k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It works with TypeScript. The repository describes itself as: eXeLearning is an AGPL-licensed free/libre tool to create and publish open educational resources. The licence is AGPL-3.0.

Example prompts

  • “/idevice”

What it can do on your machine

Read from SKILL.md and the folder at commit 319b5d2. 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

    Shell commands in SKILL.md call:

    • bun

    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

Idevice loads about 1.7k tokens when it runs. Until then it costs about 36 tokens; SKILL.md has 677 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~36
When it runs · the whole SKILL.md, loaded when a task matches
~1.7k

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 exelearning/exelearning at commit 319b5d2, republished under its AGPL-3.0 licence (© exelearning). 677 words, ~1,687 tokens.

Download SKILL.mdSave it as .claude/skills/idevice/SKILL.md (or your agent's skills folder).
name
idevice
description
Create or modify an eXeLearning iDevice, including edition/export data, legacy compatibility, assets, TypeScript bundles and teardown.

iDevice development

Read the target under public/files/perm/idevices/base/, its config.xml, edition/export scripts, colocated tests, and public/app/workarea/idevices/. Use the nearest comparable iDevice, not a generic scaffold. checklist, rubric and geogebra-activity show legacy patterns; slide/src/ shows a bundled TypeScript editor already in main (scripts/build-slide-editor.ts). Do not assume other branches' migrations landed.

Trace the complete lifecycle

  • Follow config discovery → edition initialization → validation/save → persisted HTML/data → reload → preview/export initialization → teardown. Preserve the actual $exeDevice/runtime globals and hooks. Methods vary (loadPreviousValues, updateFieldGame, etc.); there is no universal loadData() contract.
  • The test must round-trip every changed field, including empty/default values and a real legacy fixture. If stored data uses escaped/encrypted JSON or versioned attributes, keep its existing encoding contract; do not prescribe escape() for a new format or decode historical data twice.
  • Preserve IDs, config metadata, icon references, edition/export resource names and library registration. Prevent repeated .exe-text wrappers on re-import through the existing normalization path.
  • Test exported behavior independently of the workarea's globals and DOM. Multiple instances on one page need isolated IDs/state; preview and the exported ZIP must include all required resources.
  • Dispose listeners, timers, media/canvas resources and object URLs through this.$lifecycle (below); TinyMCE editors in the form are disposed centrally. Check switching devices and leaving edition; do not overwrite another device's global or discard unsaved work during a remote Yjs refresh.
  • Use _() for editor controls and c_() for learning content, preserving the project's content locale. Retain keyboard controls, labels, focus and feedback in edition and exported activities.

Edition lifecycle: own what you create

An editor is opened and closed many times in one session. Anything the edition creates that can outlive its form — a timer, a handler on document, a player, a pending read — must be owned by the edition's lifecycle, or it keeps running after the editor is gone and, worse, can end up driving the next iDevice the user opens.

IdeviceNode gives every edition an EditionLifecycle (public/app/workarea/project/idevices/content/editionLifecycle.js) before calling init(), as this.$lifecycle. Register through it and teardown is automatic. See ADR-2293-01.

javascript
init: function (element, data) {
    // Timers: cancelled on close. `this` is this edition instance.
    this.$lifecycle.setInterval(function () { this.refreshPreview(); }, 1000);

    // Handlers on shared targets: removed on close, by a namespace unique to
    // this edition, so unrelated handlers on document/window are untouched.
    this.$lifecycle.on(document, 'keydown', this.onKeyDown);

    // Native listeners: removed through the lifecycle's AbortSignal.
    this.$lifecycle.addEventListener(window, 'resize', this.onResize);

    // Anything with its own cleanup API — name the method, never assume one.
    this.$lifecycle.ownInstance(player, 'destroy');   // YouTube player
    this.$lifecycle.ownInstance(observer, 'disconnect');
    this.$lifecycle.ownMedia(audioElement);           // pause + release the stream
    this.$lifecycle.own(() => URL.revokeObjectURL(blobUrl));
    this.$lifecycle.ownFileReader(reader);            // abort an in-flight read
    this.$lifecycle.own(() => widget.teardown());     // anything else

    // Abortable requests, and file reads that always settle.
    fetch(url, { signal: this.$lifecycle.signal });
    const text = await this.$lifecycle.readFile(file, 'readAsText');
}

A resource the edition rebuilds on every user action — the audio preview, say — must name a slot, or each action leaves one more live resource and one more disposer behind until the editor closes. Owning into a slot releases whatever that slot held:

javascript
this.$lifecycle.ownMedia(new Audio(url), 'previewAudio');

Wrapping a FileReader in a promise by hand is the one case the primitives get wrong on their own: abort() fires no error event and a bound loadend no-ops, so the promise never settles and the caller awaits it for the lifetime of the page. Use readFile(file, method), which rejects with an AbortError on teardown exactly as an aborted fetch() does.

Show full SKILL.md (246 more words)Show less

Never dereference the global inside a deferred callback. When it finally runs, $exeDevice usually holds a valid different device, so a ?. guard passes and the callback edits the wrong iDevice:

javascript
setTimeout(() => $exeDevice?.refresh(), 500);                    // wrong
this.$lifecycle.setTimeout(function () { this.refresh(); }, 500); // right

Arrow callbacks ignore this, so capture the instance instead: const self = this;. A lexical capture is bound to the instance that made it, which gives the same guarantee.

Form-local handlers ($('#myFormField').on('click', ...)) need nothing: the centralized teardown unbinds the edition form through jQuery, which removes them. TinyMCE editors inside the form are also disposed centrally — do not add your own.

own() disposers run last, in reverse order. If something must be cleaned up before them, add the optional hook instead — it runs first, while the instance and its DOM are both still alive, and at most once:

javascript
destroyEdition: function () { /* ... */ },

Gotcha: a raw setInterval, a document handler or a player left running after close keeps firing, and can drive the next iDevice the user opens. Register them through this.$lifecycle. Issues #2271, #2293.

TypeScript/bundled devices

Edit the source under the device's src/ and run its registered build (slide: bun run bundle:slide-editor). Review source, tests and generated output together. Do not hand-edit the generated editor bundle or force legacy devices into a TypeScript scaffold as part of an unrelated fix.

Run the target's existing Vitest .test.js files, including source-import tests for TypeScript components. Add/update an E2E flow that creates, edits, saves, reloads and previews the device; use exporter tests for ZIP/resource changes and the collaboration fixture for remote-edit interactions. Finish with verify-change.

© exelearning, AGPL-3.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .agents/skills/idevice of exelearning/exelearning.

Open the folder on GitHubat commit 319b5d2

Compare with similar skills

Idevice 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.

Idevice compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Idevice this skillexelearning/exelearning145—~1.7kAutomated safety check: PassAGPL-3.0
MCP Server Builderanthropics/skills180k62 repos~2.3kAutomated safety check: PassApache-2.0
Web Artifacts Builderanthropics/skills180k40 repos~769Automated safety check: PassApache-2.0
Typescript Advanced Typesrolling-scopes/rsschool-app10k24 repos~4.2kAutomated safety check: PassMPL-2.0
Code Review ChecklistshareAI-lab/learn-claude-code78k5 repos~1.1kAutomated safety check: PassMIT
MCP Server BuildershareAI-lab/learn-claude-code78k5 repos~1.2kAutomated safety check: PassMIT

Similar skills

  • MCP Server Builder

    anthropics/skills

    Official

    Guides the design and implementation of Model Context Protocol servers in TypeScript or Python, from tool naming and error messages to evaluation.

    180k GitHub starsUsed in 62 repos~2.3k tokens
    Agent WorkflowsAuto-check passed
  • Web Artifacts Builder

    anthropics/skills

    Official

    Builds multi-component claude.ai HTML artifacts as a small React, TypeScript and Tailwind project, then bundles it into one shareable HTML file.

    180k GitHub starsUsed in 40 repos~769 tokens
    Frontend & DesignAuto-check passed
  • Typescript Advanced Types

    rolling-scopes/rsschool-app

    Master TypeScript's advanced type system including generics, conditional types, mapped types, template literals, and utility types for building type-safe applications.

    10k GitHub starsUsed in 24 repos~4.2k tokens
    DevelopmentAuto-check passed
  • Code Review Checklist

    shareAI-lab/learn-claude-code

    Reviews code against a five-part checklist covering security, correctness, performance, maintainability and testing, and reports findings in a fixed format.

    78k GitHub starsUsed in 5 repos~1.1k tokens
    DevelopmentAuto-check passed
  • MCP Server Builder

    shareAI-lab/learn-claude-code

    Walks through building MCP servers in Python or TypeScript that expose tools, resources and prompts to Claude, with templates, registration and testing.

    78k GitHub starsUsed in 5 repos~1.2k tokens
    Agent WorkflowsAuto-check passed
  • Nx Import

    nrwl/nx

    Import, merge, or combine repositories into an Nx workspace using nx import.

    29k GitHub starsUsed in 5 repos~3.5k tokens
    DevelopmentAuto-check passed

More from exelearning/exelearning

All 17 skills in this repo
  • E2E Test

    exelearning/exelearning

    Write or debug repository Playwright specs, including isolated projects, preview frames and multi-user fixtures.

    145 GitHub stars~1.1k tokensUpdated today
    Auto-check passed
  • Frontend Module

    exelearning/exelearning

    Change vanilla JavaScript UI modules in public/app/, including DOM state, events and localization.

    145 GitHub stars~867 tokensUpdated today
    Auto-check passed
  • I18n

    exelearning/exelearning

    Add or change source-level localized strings using (), c() or the Nunjucks trans filter; not catalog extraction.

    145 GitHub stars~812 tokensUpdated today
    Auto-check passed
  • Verify Change

    exelearning/exelearning

    Choose and run the appropriate eXeLearning checks for a diff, including the final code gates or guidance-only validation.

    145 GitHub stars~676 tokensUpdated today
    Auto-check passed
  • Websocket Yjs

    exelearning/exelearning

    Change eXeLearning shared document mutations, WebSocket rooms, snapshots, reconnects or concurrent editing behavior.

    145 GitHub stars~759 tokensUpdated today
    Auto-check passed
  • API V1

    exelearning/exelearning

    Add or change external REST API v1 endpoints and their documented authentication/authorization contracts.

    145 GitHub stars~427 tokensUpdated today
    Auto-check passed

Works with

Questions about Idevice

What does Idevice do?

Create or modify an eXeLearning iDevice, including edition/export data, legacy compatibility, assets, TypeScript bundles and teardown. Idevice is an agent skill from exelearning/exelearning. Create or modify an eXeLearning iDevice, including edition/export data, legacy compatibility, assets, TypeScript bundles and teardown.

How do I install Idevice in Claude Code?

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

How do I install Idevice in Codex?

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

Can I use Idevice 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 exelearning/exelearning --skill idevice -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/idevice, .gemini/skills/idevice, .github/skills/idevice and .opencode/skills/idevice in your project.

What does Idevice need to run?

Going by SKILL.md and its folder, Idevice needs the command-line tools its instructions call (bun).

Does Idevice 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 Idevice 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 Idevice use?

Idevice is published under the AGPL-3.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Idevice use?

About 1.7k tokens (SKILL.md is roughly 6.7k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Idevice?

Skills that share tags, products or a category with Idevice: MCP Server Builder (anthropics/skills, 180k stars), Web Artifacts Builder (anthropics/skills, 180k stars), Typescript Advanced Types (rolling-scopes/rsschool-app, 10k stars) and Code Review Checklist (shareAI-lab/learn-claude-code, 78k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Idevice?

exelearning (a GitHub organization) maintains it in exelearning/exelearning, which has 145 GitHub stars. The repository holds 17 skills in this directory. The repository was last updated on October 7, 2026.

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