yokami is an independent software publisher distributing its work through GitHub, with a catalog currently centered on a single product: P-ai. P-ai is a ready-to-use desktop AI assistant designed to operate autonomously over long-running tasks rather than handling only short, one-off prompts. Its feature set targets users who want an assistant that can grow and adapt over time, with built-in memory that retains context across sessions, support for multiple agents that can be orchestrated for complex workflows, and tool review capabilities that let users supervise what the assistant does with external utilities. Integration with MCP, the Model Context Protocol, allows the assistant to connect with a broad ecosystem of compatible tools and data sources, extending its usefulness beyond a fixed set of functions. The product also emphasizes a high-concurrency workspace, meaning it can manage many tasks or agents in parallel, which suits power users, developers, and researchers running several automated processes at once. Typical use cases for software in this category include automating repetitive desktop workflows, delegating multi-step research or coding tasks to agents, maintaining a persistent personal knowledge assistant that remembers prior work, and coordinating tool-driven pipelines without constant manual oversight. As a desktop application, P-ai fits scenarios where users prefer local, always-available assistance tied directly to their working environment rather than a purely cloud-based chat interface. With its focus on self-growing behavior, memory, and agent orchestration, yokami positions P-ai within the emerging category of agentic AI desktop software aimed at sustained, semi-autonomous productivity rather than simple question answering.
A ready-to-use self-growing desktop AI assistant for long-running tasks, memory, agents, tool reviews, MCP, and high-concurrency workspace automation.
Details