repo-alignment-scanlisted

# Repo Alignment Scan ## Purpose Digest the requirement, explore the actual repository (conventions, stack, existing patterns), and surface where the requirement is misaligned with what already exists — before Think finishes framing it. Prevents a plan from being built on an assumed architecture that the real repo already contradicts. This is a power skill, not a lifecycle phase. It is passive/scored: the agent evaluates the trigger predicate against recorded signals and decides whether to run it, recording that decision. ## Invocation Context Use this skill when the recorded `skill_scoring` trigger for `repo-alignment-scan` evaluates true: `complexity_score >= 40 OR new_surface OR task_class != trivial`. Do not invoke it for Trivial work with no new surface and a low complexity score — the trigger predicate itself should evaluate false, and Think should record a `skipped` decision with that reason rather than running the skill anyway. ## What To Load **Foundation** (confirm in context; load if not already present): - Root `AGENTS.md` - `workflow/router.md` - `workflow/lifecycle.md` - `workflow/rules.md` **Minimum for invocation**: - This file - `references/output-schema.md` **Before starting work**: - `references/scan-method.md` — what "the actual repo" means to inspect and how to compare it against the requirement - `workflow/config/repo-profile.yaml` — protected paths, public contracts, generated outputs - The requirement as framed so far (brief draft or user req