Pi Network announced yesterday another expansion of SoloHost, adding OpenClaw and Atlassian MCP Server as featured applications available through the native Pi Desktop.

The two new additions were initially introduced alongside the recent Node 0.6.2 update, which we reported a few weeks ago, but are not being formally highlighted by the team as examples of how they intend to expand the utility of their Nodes beyond simply supporting the blockchain.

AI Push Expanded

The official blog post from the team highlighted the more broadly applicable addition called OpenClaw. It’s a locally run AI agent capable of assisting users with various tasks and operating using either a locally hosted AI model or external ones such as ChatGPT and Claude.

Its memory is stored locally on the user’s computer in either configuration, although requests sent to cloud models are still processed by the external providers.

The team explained that running OpenClaw through SoloHost places the agent inside a container that restricts its default access to unrelated files and resources on the user’s computer. Pi Desktop handles much of the technical setup automatically, reducing the need for Pioneers to manually configure servers, Docker environments, and other infrastructure.

OpenClaw joins Hermes, another local AI agent already available through SoloHost.

Second Addition

The Core Team outlined the second major change, Atlassian MCP Server, which targets a more specialized audience. It allows developers and professional teams to run their own local MCP server and connect compatible AI tools with Jira.

Pi Network said both of these new additions aim to transform its infrastructure into a practical computing platform. Recall that SoloHost saw the light of day on Pi2Day (June 28) and enables third-party devs to publish self-hosted applications that Pioneers can run through Pi Desktop.

In theory, this allows developers to access Pi’s network of over 420,000 claimed node operators, while providing Pioneers with additional uses for the computing resources they already operate.

Aside from the aforementioned additions, the Core Team recently introduced new pricing for its App Studio model, moving away from subsidized rates to reflect the actual AI service costs. The change went live on August 24.

Source: CryptoPotato