Task: Strip opinions from loop-harness.md, leaving mechanics #12

Goal

Strip from docs/loopwright/loop-harness.md:44-77 the rules that are core
opinions, leaving behind only harness mechanics. Today that file carries eight
bolded "Rules of the loop", several of which assert the same things the new doc
asserts — two lists of rules is two sources of truth, and the one nobody updates
starts lying.

This task does not write any principle. principles.md is the sole author of
P1–P12; its own task writes them in full. The change here is one-directional:
delete the opinion from loop-harness.md and replace it with a reference to the
principle that now owns it.

It also reconciles the one place where the new doc changes a rule rather than
describing it: line 66 states that merging is always the human's decision,
absolutely, while P8 makes it opt-in-automatable.

Acceptance criteria

• All eight rules at loop-harness.md:44-77 are handled per this fixed
  classification, with no rule left undecided:

  | Lines | Rule | Verdict |
  |---|---|---|
  | 46-49 | Base branch | mechanics — stays |
  | 50-54 | Closing keywords | mechanics — stays |
  | 55-57 | Draft until ready | mechanics — stays |
  | 58-60 | The gate is the only source of truth for "done" | opinion **P2** — deleted, replaced by a reference |
  | 61-65 | State lives in artifacts, not sessions | opinion **P1** — deleted, replaced by a reference |
  | 66 | Merging is always the human's decision | opinion **P8** — deleted, replaced by a reference |
  | 67-72 | Architecture is scaffolding, not a spec | mechanics — stays |
  | 73-77 | `boundary-reviewer` is the first thing to cut | mechanics — stays |
  

• No rule that is an opinion remains stated in loop-harness.md — the file
  references the principle by its id instead of repeating it, so no rule
  exists in both files and no principle text is authored here.

• The merge rule at loop-harness.md:66 is replaced by a bare reference to
  P8 — not by a restatement of it. loop-harness.md may still describe that
  a merge step exists at the end of the flow, which is mechanics; it must not
  state who holds the authority or under what condition, which is P8's.
  Concretely: "See P8" and not "merging is the human's by default, automating
  it is an opt-in" — the second sentence is principle text, and authoring it
  here is what recreates the two-sources-of-truth problem this task exists to
  remove.

• What remains in loop-harness.md still reads coherently on its own as the
  execution half — base branch, closing keywords, draft-until-ready, the flow
  diagram and the Pieces table are intact.

• docs/loopwright/quality-gate.md:7-9 ("The one rule") is left untouched:
  "only quality-gate.mjs can fail the build" is CI mechanics, not an
  opinion, and P2 does not replace it.

Pointers into the RFC

• Scope, second bullet: what moves and what stays.
• "Why the contested branches settled where they did" → the P2/P8 split, and
  the explicit note that the quality-gate.mjs rule is mechanics living in a
  different doc.
• Risks → P8 is the one non-descriptive line in a descriptive document; the
  reconciliation of line 66 is where that shows up in the repo.

Blocked by

#11 — the destination has to exist before anything moves into
it, or the move produces a dangling reference.