align
align — AI-initiated alignment with pre-selected defaults
The sponsor decides better by selecting than by composing. The agent does the thinking — frames the question, enumerates the realistic options, ranks them, and pre-selects the best — so the sponsor only has to confirm. Every tunan skill routes its decision points through this single protocol instead of inventing its own way to ask.
When this runs
- Direct: the sponsor types
/tunan:align <topic>. - Indirect (primary): any other tunan skill hits a decision point, builds the question per this protocol, collects the sponsor's choice, then returns to its own flow.
Core rule
The agent never hands an open-ended question back to the sponsor. Every decision point becomes: a one-line question plus at least 3 ranked recommendations, with the single best one pre-selected as the default. "Up to you" / "either works" / "what do you prefer" are forbidden — surfacing options without a recommendation defeats the purpose.
Interaction Method
Default to the platform's blocking question tool: AskUserQuestion in Claude Code (call ToolSearch with select:AskUserQuestion first if its schema isn't loaded), request_user_input in Codex, ask_user in Gemini, ask_user in Pi (requires the pi-ask-user extension). Fall back to a numbered list in chat only when no blocking tool exists in the harness or the call errors (e.g., Codex edit modes) — not because a schema load is required. Never silently skip the question.