architecture-lead

This skill should be used when a specific design choice has been made or is being considered during a Gate 2 conversation, or the user says "log a decision", "write an ADR", "capture this choice", or "/architecture-decision-record". It captures the choice in a standard ADR format and stores it under outputs/adrs/ as the engagement's internal audit trail. For structural decisions, it dispatches to the cartographer for a supporting diagram.

This skill should be used once Gate 2 is complete and the user says "write the brief", "draft the client document", "produce the decision brief", or "/decision-brief". It dispatches to the cartographer for the brief's diagram set, drafts the six-section client-facing decision brief with those diagrams stitched into Section 2 — the key output artifact of the engagement — and writes it to both outputs/decision-brief.md and outputs/decision-brief.docx.

This skill should be used when the user pastes raw, unformatted discovery notes from a client session, says "structure these notes", "run discovery", "requirements gathering", or "/requirements-gathering", or is at the start of a new architecture-lead engagement that has not yet reached Gate 1. It ingests notes/ (or pasted text) plus reference/ compliance docs, distills a plain-language problem statement first, structures the rest against a discovery checklist, runs the skeptic/practical-planner negotiation, and produces a Gate 1 findings summary.