What is Product Brain?
Product Brain is a governed knowledge graph for AI-assisted product work. It gives your agents a shared source of truth: decisions linked to the tensions that caused them, standards linked to the work they constrain, insights linked to the work packages they inform.What is the Chain?
The Chain is the typed, versioned graph inside Product Brain. It holds decisions, tensions, standards, business rules, architecture entries, work packages, and insights. Each entry has an ID, relationships, and enough structure for agents to query it through MCP before they act.How does it work with Cursor, Claude, or Codex?
Product Brain exposes the Chain through MCP. Add the MCP configuration to your tool, then you ask in plain language and your agent callsorient, entries, capture, and session tools—without you re-pasting project history every time.
Who is it for?
Product Brain is for builders and product-led teams doing serious work with AI agents. It is useful when the problem is not missing notes, but scattered context and uncertainty about which decisions are still true.What is the Loop?
The Loop is the working rhythm around the Chain: orient before meaningful work, build with relevant context loaded, capture durable decisions and tensions as they appear, then close the session so the next person or agent starts grounded. See The Loop.Brain Chat vs MCP?
Brain Chat is the in-app workspace chat (Brain Chat). MCP is for external agents (Cursor, Claude, ChatGPT, …). They read the same Chain but run in different hosts.Import vs paste?
UI import walks batches through review before commit (Import & review).pb ingest is CLI-first bulk ingestion for repos—ask your agent when you have filesystem access.