owner-meeting

To keep client meetings focused on requirements alignment, enforce a discussion mode where the AI summarizes known context, asks clarifying questions, and waits for explicit execution triggers before writing code or splitting tasks. This prevents premature execution and ensures alignment.

The best way to document architecture decisions during product discovery meetings is to verbally summarize the discussion, confirm with the client, and then record the conclusions directly into `.boss/architecture` and `.boss/requirements` directories to maintain a formal decision trail.

To perform a coverage check for complete product linkages, expand mentioned anchors into full paths, state coverage, security baselines, and data lifecycles, then present these expansions to the client in a single sentence for immediate confirmation to complete the product net.

Recorded decisions should be scanned for compatibility right before finalizing new architecture documentation. You must resolve conflicts with old decisions during the meeting and define migration strategies before saving updates to the `.boss/architecture` directory.

Yes, you can use this meeting facilitation approach for knowledge Q&A without triggering execution. When the client asks informational questions, the AI answers using layered explanations and can optionally propose saving valuable insights to the wiki without starting any execution.

The transition from discussion mode to task execution is triggered exclusively by explicit client directives such as "start execution" or "write code." Until the customer explicitly commands it, the session remains in continuous questioning and alignment mode.