Metis Board

Most project management tools were designed for humans updating tickets by hand. AI agents can talk around those systems, summarize them, or automate small pieces of them, but the underlying project state is still not agent-native. Metis Board removes that friction by treating human operators and AI agents as participants in the same governed work system: the same work items, the same permissions, the same memory, the same audit trail, and the same delivery contracts. Metis Board is being built for teams that expect AI agents to participate in delivery, not sit beside it. The product combines familiar project surfaces - workspaces, projects, work items, boards, sprints, releases, dependencies, documents, activity, audit, and lightweight reporting - with an MCP-accessible action layer, shared agent memory, and methodology-aware governance. Legacy project management tools can add AI summaries and assistants, but the AI is usually bolted onto a system designed for manual ticket work. Metis Board starts from the opposite assumption: agents need protocol-grade access, scoped permissions, durable memory, and visible provenance if they are going to participate in real delivery. Inside the Active Twist constellation, Metis Board is the work substrate: the place where cards, acceptance criteria, phase gates, halt triggers, delivery contracts, decisions, and agent memory become operational state. It pairs naturally with MCD, AmphionAgent, JSONM design-system portability, and MaMC methodology templates because it is built around the same assumption: AI work needs structure to stay useful. Metis Board is in active build and close to GTM. Until its standalone product site ships, this Active Twist spoke is the interim home for launch interest and product-fit conversations.