> For the complete documentation index, see [llms.txt](https://docs.miuro.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.miuro.ai/miuro-claw/miuro-claw-roadmap.md). # Miuro Claw Roadmap MIURO Claw will be developed through phased iteration. The priority is to complete the core closed loop first, then gradually improve edge capabilities and security hardening. Each phase will be managed through an OpenSpec Change to ensure the architecture remains stable and controllable. ### Phase 1: Cloud Thin Layer + Full Migration #### Goal: Quickly validate cloud agent capabilities and establish the foundation for a dual-engine abstraction. #### Main Scope * Develop `ThinClawEngine`, supporting synchronous conversation, asynchronous songwriting cards and a read-only skill subset * Implement account-level memory through `claw-account-memory` * Fully migrate the existing cloud OpenClaw to ThinRunner, without parallel grayscale coexistence * Lock down the `clawrunner.Runner` abstraction and shared foundation boundaries #### User-Visible Value * Cloud Claw becomes faster and more lightweight * Supports basic conversational songwriting and simple tool calls #### Estimated Stage: Change 1, highest priority *** ### Phase 2: Device Closed Loop #### Goal: Enable the user’s device to become a truly independent Claw instance. #### Main Scope * Develop `OpenClawRunner` with a single-tenant architecture * Implement one-click factory pre-installation and a zero-screen activation flow on first boot * Introduce device-side software keys and per-request signing * Add persistent MQTT signaling and on-demand wake-up warm-up * Complete the full message routing loop between device and cloud #### User-Visible Value * Users can truly use device-side Claw after purchasing MiuroBox * Supports collaboration between local edge capabilities and cloud models * Enables remote wake-up from web or app to start creation on the device #### Estimated Stage: Change 2 *** ### Phase 3: Edge Audio AI + Device-Exclusive Benefits #### Goal: Strengthen local audio processing capabilities on the device side and define hardware ownership benefits clearly. #### Main Scope * Launch Edge Audio AI, Phase A, using sherpa-onnx for local noise reduction, ASR and TTS * Implement “zero-upload recording studio”, “24/7 inspiration guardian” and “offline live mode” * Connect the device key benefit axis, including local downloads, offline music library, automation and full skill access * Clearly define the commercial boundary between permanently owned local functions and on-demand cloud capabilities #### User-Visible Value * Users can complete high-quality audio processing locally without uploading private data * Partial music creation workflows can still run offline * Hardware buyers receive clear device-exclusive benefits #### Estimated: StageChange 3 *** ### Phase 4: Workflow Orchestration and Deeper Skill Ecosystem #### Goal: Turn MIURO Claw into a true full-state agent with complex workflow orchestration capabilities. #### Main Scope * Implement the “hum one line and generate a song” closed loop * Support full-chain orchestration through one sentence: lyrics → composition → arrangement → mastering → library storage * Introduce a provenance mechanism to record creative contribution traces * Open the skill ecosystem and pre-install popular skills * Support 24/7 automated task execution #### User-Visible Value * Upgrades from an “assistant tool” into a complete music studio * Users can drive complex creative workflows through natural language * Significantly improves creative efficiency and consistency #### Estimated Stage: Change 4 ### Phase 5: Security Hardening and Performance Optimization #### Goal: Improve system security and device-side performance to support larger-scale user adoption. #### Main Scope * Add local encryption at rest and optional encrypted backup recovery for personality and memory data * Introduce secure element hardware keys, such as ATECC608B / SE050, in the P2 stage * Accelerate selected TTS capabilities with NPU, Phase B * Explore OP-TEE and Secure Boot in the P3 stage * Further optimize device communication stability and relay capacity #### User-Visible Value * Higher data security and stronger privacy protection * Better device-side performance * Preparation for future large-scale commercial deployment and enterprise scenarios *** | Phase | Core Goal | Key Capabilities Delivered | User Value Level | Estimated Priority | | ----- | ----------------------------------------------- | --------------------------------------------------------------- | ------------------------------ | ------------------ | | 1 | Cloud thin layer + full migration | ThinRunner + account-level memory | Basic usability | Highest | | 2 | Device closed loop | OpenClawRunner + MQTT communication | Device-side usability | High | | 3 | Edge audio + exclusive benefits | Local noise reduction / ASR / TTS + hardware ownership benefits | Privacy and offline capability | Medium | | 4 | Full workflow and skill ecosystem | Agent orchestration + automation | Complete music studio | Medium | | 5 | Security hardening and performance optimization | Hardware keys + NPU acceleration + encrypted backup | Enterprise-grade reliability | Lower | --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.miuro.ai/miuro-claw/miuro-claw-roadmap.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.