Your first useful plan / Votre premier plan utile
Visit the welcoming homepage · Make your first plan · Commencer en français
No technical experience, account, wallet, payment, installation or AI service key is needed. This page runs a small planning calculation on your device. It does not perform the tasks for you.
Find your way from the homepage
The homepage starts with one clear invitation to make a first plan. An example shows what you will receive, three short steps explain the process, and plain-language questions explain cost, privacy and what happens next. Choose Français or Larger text at the top before you begin. The first-plan link carries those choices with you; there is no need to select them again. The make-plan buttons open the task choices directly. The general Start here navigation link keeps the short introduction.
You can also open the French experience with larger text directly. Preferences are carried in the link for this visit, without an account or saved browser settings. The advanced workspaces, research, documentation and original demonstrations remain farther down the homepage for people who want to explore them.
Three easy steps
- Choose. Enter your available minutes. Keep the four sample tasks or change their names, durations and importance points. Use whole numbers: 1–480 minutes and 1–10 importance points.
- Review. Select Find my plan. Read the selected and omitted tasks and the minutes remaining. The planner compares all 16 combinations and maximizes the importance points within your time. Tasks must be independent and fit in full; it does not account for dependencies or interruptions. Check your assumptions and tick the review box, or select Change my choices.
- Keep. Select Continue to save, then Download my plan. The text file is easy to read. You can also select Read my plan here or Print or save as PDF. Save evidence downloads JSON with the exact inputs, task names and calculated result. A checked box records your review; it is not independent verification or permission to act.
The planner does not send task entries to a server, use analytics, save them in browser storage, connect a wallet or call an AI provider. Hosting still receives ordinary page and asset requests. Reloading clears your entries; keep the downloaded file before leaving. Other workspaces have their own storage and recovery behavior, described in the browser guide.
Make yourself comfortable
- Use Larger text, browser zoom, or Français at the top. Homepage language and text-size choices follow the first-plan link; they are for this visit and can be changed at any time.
- Use Tab and Shift+Tab to move between controls, Enter to activate a button, and Space for a checkbox.
- There is no countdown or automatic progression. Changing a plan clears the previous review and result.
- Error messages name the field to fix. Select the linked message to reach that field.
- If no task fits, increase available time or reconsider the estimates. An empty selection is a valid result.
- Reading controls appear when the planner is ready. If JavaScript is unavailable or a required module cannot load, the page keeps readable guidance instead of showing controls that cannot work.
- If the planner does not load, follow the on-page guidance or reload online in an updated browser.
- Downloads go to the browser’s Downloads folder; on a phone, use the browser’s download menu or Files app.
Continue when you are ready
| What you want | Where to go | What to expect |
|---|---|---|
| Compare a more detailed decision | Decision Studio | Browser calculations for resources, schedules and forecasts |
| Explore agent improvement | SUCCESSOR Ω | Bilingual bounded mission experiment with explicit evidence limits |
| Explore the project | All workspaces | Current tools and preserved historical demonstrations |
| Install an agent | Installation guide | Python 3.11–3.13, matching release files, local setup and private state |
| Rehearse contract deployment | Hardhat guide | Developer-only, disposable local EVM; no real funds or network deployment |
Français
Ouvrir la page d’accueil en français ou commencer directement avec un texte plus grand. La page d’accueil présente un exemple de plan, trois étapes et des réponses aux questions courantes. Vos choix de langue et de taille du texte vous accompagnent au début du parcours. Les boutons pour créer un plan ouvrent directement le choix des tâches; le lien général Commencer garde l’introduction. Aucun compte, portefeuille, paiement, téléchargement de modèle ni installation n’est nécessaire.
- Choisir. Indiquez le temps disponible et conservez ou modifiez les quatre tâches proposées. Utilisez des nombres entiers : 1 à 480 minutes et 1 à 10 points d’importance.
- Vérifier. Sélectionnez Trouver mon plan. Vérifiez les tâches sélectionnées, les tâches omises et le temps restant. Le calcul compare les 16 combinaisons; les tâches sont indépendantes et indivisibles. Cochez la case de vérification ou choisissez Modifier mes choix.
- Garder. Sélectionnez Continuer pour enregistrer, puis Télécharger mon plan. Vous pouvez aussi choisir Lire mon plan ici ou Imprimer ou enregistrer en PDF. Le fichier texte est lisible; le fichier JSON contient les données exactes et le résultat.
Les données restent sur la page et ne sont pas enregistrées par ce planificateur. Rechargez la page pour les effacer; téléchargez votre résultat pour le conserver. Le site reçoit les requêtes ordinaires nécessaires pour afficher la page. Aucun achat, publication ni travail externe n’est effectué. Le bouton Texte plus grand, le zoom et la navigation au clavier sont disponibles, sans minuterie. Les réglages de lecture s’affichent quand le planificateur est prêt. Si le chargement échoue ou si JavaScript est désactivé, les indications restent lisibles sans afficher de commandes inutilisables. Les espaces avancés ont leurs propres règles de stockage et peuvent demander une configuration technique.
Maintained scope and scale
The first-use experience is static, small and independent of a central inference service. The four-task calculation has a fixed 16-combination bound. This reduces per-user server work and offers a practical entry point for a broad audience. It is not evidence of billion-user hosting capacity or a public, multi-tenant agent service. Capacity planning, regional delivery, abuse controls for future APIs, additional browser and assistive-technology testing, and real-user usability studies remain separate work. See release readiness for the supported deployment scope.