TECNO EllaClaw AI Agent Gets Cross-App Intelligence, 40+ Smart Skills in Closed Beta

TECNO's EllaClaw AI Agent Can Now Shop, Book Rides, and Clean Up Your Phone

TECNO EllaClaw just got more capable. The AI agent, built on top of TECNO's Ella assistant, can now act inside other apps on your behalf. It can also handle some of the day-to-day upkeep on your phone itself. EllaClaw is still labeled Beta, running as a cloud-based agent in the background rather than a local, on-device feature. It's part of TECNO's broader Practical AI push. The goal is making AI tools genuinely useful for people in markets where a phone is often someone's main computer. "Our goal is to create AI that helps simplify everyday tasks," said TECNO General Manager Jack Guo. He also pointed to keeping users in control of what the agent does. Phone caretaker duties On the maintenance side, TECNO EllaClaw can free up RAM and CPU through a feature called Smart CleanUp Boost. It can also flag power-hungry apps draining your battery, and cool things down during heavy tasks like gaming. There's also a data usage tracker aimed at people watching their mobile data closely. Major changes still need your confirmation before EllaClaw follows through. Acting inside other apps The bigger shift is cross-app intelligence. EllaClaw can book a ride through natural conversation and check in on connected smart home devices. It can also help find products inside shopping apps like Lazada. TECNO says it does this by reading app screens the way a person would, rather than relying on hidden API access. That means you can watch and follow what it's doing, step by step. TECNO EllaClaw is still in closed testing, with no word yet on when, or if, any of this reaches a public release.

TECNO EllaClaw AI agent gets cross-app intelligence, persistent memory, and 40+ Smart Skills

TECNO has showcased the next stage of evolution of EllaClaw, its beta-stage mobile AI agent designed to bring practical agentic AI into everyday mobile use. TECNO positions EllaClaw under its Practical AI vision, shifting from reactive responses to proactive, task-oriented intelligence. The system combines device-level management, cross-app automation, and contextual assistance while maintaining strict permission-based controls so users approve key actions. It extends optimisation across mobile data, battery life, and storage, enabling the device to handle more complex tasks on behalf of the user. System-level optimisation with Smart Skills EllaClaw includes more than 40 Smart Skills for device management and performance optimisation. Key functions include: * Smart CleanUp Boost to reduce lag and free system resources * Smart Power Drain Check to identify battery-consuming apps * Instant Cool-down Relief to manage background activity during heavy usage * Smart Data Guardian (India-specific) for monitoring data usage patterns The system follows a confirmation-first model, requiring user approval before executing system-level actions. Context-aware assistance with persistent memory EllaClaw uses persistent memory to adapt to user behaviour and provide contextual assistance across daily tasks. It supports: * Morning briefings with calendar, weather, itinerary, and news updates * Trip planning with ride setup, alarms, and schedule organisation * Reminders for tasks and communication follow-ups * Real-time weather tracking and alerts * Context-based reminders for personal connections and follow-ups Cross-app intelligence with controlled automation EllaClaw can operate across multiple applications using opt-in permissions and non-intrusive GUI-based navigation, keeping actions visible and user-controlled. It can interact with: * Shopping applications * Transport and ride-hailing services * Food delivery platforms * Smart home systems * Messaging applications Use cases include one-sentence ride booking, smart home monitoring, and shopping assistance across e-commerce platforms. Availability EllaClaw is currently in internal testing and closed beta, with broader availability details yet to be announced, the company says.