Add local GUI and secure Claude Input profile generator

[thannous]

Résumé

• ajoute un configurateur local Vite lancé par npm run configure pour personnaliser les contrôles sûrs du Codex Micro ;
• génère un profile Work Louder Input personnel depuis l’export officiel de l’utilisateur, sans publier ni modifier cet export ;
• préserve le layer natif à l’index 0, les autres layers, leur ordre et le lien AppSense Claude existant ;
• consolide les schémas, validateurs, tests, outils de sauvegarde/dry-run/rollback et la documentation de la V1 sécurisée ;
• garde le manifeste V1 hardware-observed et le contrat logique public V0 proposal-not-applied.

Parcours utilisateur

1. exporter le profile actif depuis Work Louder Input ;
2. lancer npm run configure et charger cet export dans le GUI local ;
3. vérifier l’unique layer Claude lié avec AppSense ;
4. personnaliser les quatre touches, le cadran et le joystick ;
5. télécharger Claude-macOS-profile.json ;
6. importer ce nouveau fichier avec Add New.

Le JSON logique du dépôt n’est pas présenté comme un preset universel importable.

Validation effectuée

• npm ci --ignore-scripts --no-audit --no-fund
• npm run check
  • schémas et invariants V0/V1 validés ;
  • 17 tests Node racine réussis ;
  • build Vite de production réussi ;
  • 4 tests dédiés au worker GUI réussis après le build, soit 21 tests au total ;
  • liens locaux vérifiés dans les 22 fichiers Markdown suivis par Git.
• simulation d’un runner propre sans prototype/node_modules : installation verrouillée de 66 packages sans scripts, puis suite complète réussie ;
• npm run configure -- --help
• git diff --check
• diagnostic local : Work Louder Input 0.17.3 et Claude Desktop 1.24012.9 compatibles ;
• dry-run sur un export officiel local : unique layer Claude à l’index 1, index 0 protégé, AppSense présent, aucun bloqueur ;
• génération réelle dans un dossier temporaire : layer natif strictement identique, AppSense conservé et nombre de layers inchangé.

Sécurité et réseau

Les profils sont traités localement, sans télémétrie applicative. Lors de leur première exécution, npm run configure et npm run check peuvent exécuter npm ci --ignore-scripts dans prototype/ afin de télécharger les dépendances verrouillées du GUI.

Limites

• aucun export utilisateur, identifiant AppSense personnel ou secret n’est inclus ;
• cette validation n’a pas répété le test sur clavier physique ;
• le round-trip public d’un artefact *-layer.json, la perte de focus et le comportement physique complet restent à confirmer avant toute promotion supplémentaire du statut.

[coderabbitai]

Warning

Review limit reached

@thannous, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 41 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: 33f5ede5-f1e9-48a3-8156-0d0898a3e27c

📥 Commits

Reviewing files that changed from the base of the PR and between b17b2ba and 37c3029.

⛔ Files ignored due to path filters (3)

• prototype/package-lock.json is excluded by !**/package-lock.json
• prototype/public/assets/ai-controller-claude-v1.png is excluded by !**/*.png
• prototype/public/assets/claude-glass-background-v1.png is excluded by !**/*.png

📒 Files selected for processing (27)

• .github/PULL_REQUEST_TEMPLATE.md
• .github/workflows/validate.yml
• .gitignore
• README.md
• SECURITY.md
• docs/codex-micro/configuration.md
• docs/installation.md
• package.json
• profiles/claude-shortcuts/README.md
• profiles/claude-shortcuts/macos.example.json
• profiles/claude-shortcuts/schema.json
• prototype/.npmrc
• prototype/.openai/hosting.json
• prototype/AGENTS.md
• prototype/index.html
• prototype/package.json
• prototype/scripts/prepare-sites-build.mjs
• prototype/src/App.jsx
• prototype/src/main.jsx
• prototype/src/styles.css
• prototype/tests/sites-worker.test.mjs
• prototype/vite.config.mjs
• prototype/worker/index.js
• scripts/configure.mjs
• scripts/lib/gui-dependencies.mjs
• scripts/prepare-gui.mjs
• scripts/validate-profile.mjs

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch codex/consolidate-shareable-input-v1

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 37c302956b

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Verify the AppSense target before accepting the backup

inspectInputProfile only checks that linkedAppId is a nonnegative integer and then reports appSenseLinked: true; it does not establish that the ID targets Claude Desktop. Consequently, a layer named Claude that is actually linked to another application is accepted as verified, and the generated profile preserves that incorrect link, causing these shortcuts to activate in the wrong foreground application. Validate the target application identity or require explicit confirmation before enabling generation.

AGENTS.md reference: prototype/AGENTS.md:L15-L15

Useful? React with 👍 / 👎.