Today, we launch DClaw — DGrid’s foundational personal agent layer for the open agent economy. At DGrid, we believe the next era of AI will not be defined by isolated chatbots, but by personal agents: intelligent systems that operate across your tools, learn your context over time, act on your behalf, and remain fully under your control.
Built on top of CoPaw, DClaw streamlines the creation of persistent, multi-channel, user-owned AI agents with minimal setup. It lowers the barrier to entry for the agent economy, making personal AI infrastructure accessible to individuals, developers, teams, and communities alike.
DClaw is not just another AI interface — it is a new operating model for how AI integrates into everyday life.
What Is DClaw?
DClaw is a one-click deployment engine for personal AI agents, purpose-built for the DGrid ecosystem. It enables users to launch autonomous agents that communicate across channels, retain persistent context, execute real-world tasks, and evolve into long-term digital companions.
Each DClaw instance functions as an intelligent node in the broader DGrid network, bridging personal utility with open ecosystem participation. In practice, DClaw transforms AI from a tool you visit into an agent that works where you work.
More importantly, DClaw is designed to make agent deployment radically simpler than traditional open-source frameworks such as OpenClaw. While OpenClaw gives developers a flexible base for building agents, DClaw turns that foundation into a production-ready, user-facing deployment layer:
- 1-click deployment instead of complex environment setup and manual configuration
- All-in-one DGrid API access, so users can get started without stitching together multiple external services
- Integrated model access, channels, memory, and skills in a single deployment experience
- Faster path from agent idea to live agent, for both technical and non-technical users
In that sense, OpenClaw is closer to an open framework, while DClaw is the deployment layer that makes personal agents truly usable at scale.
Native Integration: Built for How People Live and Work
Friction remains the biggest barrier to AI adoption: new apps, disjointed interfaces, unfamiliar workflows, and repetitive onboarding. DClaw eliminates this by integrating directly into the communication and productivity environments users already rely on.
DClaw supports a rapidly expanding set of channels and protocols:
- Telegram, iMessage, Matrix
- DingTalk, WeCom, WeChat
- MQTT, Twilio-based communication APIs
A single agent maintains a unified presence across work, social, and developer environments without fragmented identities or repeated setup. Instead of forcing users into yet another AI platform, DClaw brings intelligence to where coordination, communication, and action already happen.
User-Owned Memory: Trust as a Foundation
Personal agents gain value over time through accumulated experience, but long-term utility requires trustworthy memory. DClaw includes a persistent memory system designed from the ground up around user control.
DClaw remembers:
- Past conversations and interactions
- User preferences and habits
- Recurring workflows and task history
- Contextual information that improves future performance
Unlike opaque AI systems that silently collect and retain data, DClaw operates on the principle of user-owned memory. Users decide what is stored, updated, or forgotten. Personalization is not data extraction — it is an extension of user sovereignty. This is a foundational difference: in the long run, the most useful AI will not just be the most capable, but the most trustworthy.
Extensible by Design: Skills, Plugins, and Self-Improving Workflows
A powerful agent should grow with its user’s needs. DClaw features a modular skills and plugins architecture that allows agents to gain new capabilities without full rebuilds:
- Access to a community skill hub
- Support for custom skill development
- Hot-swappable skills (no agent restart required)
- AI-generated skills for dynamic workflow creation
DClaw is not just configurable — it is programmably accessible. A marketer can automate weekly campaign intelligence reports; a founder can track ecosystem signals and synthesize insights; a developer can build a deployment, monitoring, or debugging assistant. Over time, agents become increasingly specialized, not through vendor lock-in, but through continuous, user-driven capability expansion.
Built-In Access to Leading AI Models
DClaw is designed for effortless deployment, so users do not need to manage separate model APIs to get started. It includes built-in access to top-tier AI models with zero extra API key configuration:
- Claude Opus 4.6, GPT-5.4
- Kimi K2.5, Qwen 3.6 Plus, DeepSeek V3.2
DClaw intelligently routes tasks to the most suitable model based on capability, latency, and cost, while also allowing users to manually select their preferred model. This delivers flexibility without operational overhead and offers developers a fast path to production-grade agent behavior.
Get Started Today
The era of personal, user-controlled AI agents has arrived. No more app switching, no more lost context, no more surrendering data ownership to closed platforms.
With DClaw, anyone can deploy a personal agent in minutes, participate in the DGrid ecosystem, and help shape the future of the open agent economy.