diff --git a/quest/m0/README.md b/quest/m0/README.md
index 92d57161..1effb781 100644
--- a/quest/m0/README.md
+++ b/quest/m0/README.md
@@ -94,6 +94,7 @@ ownership-manifest design:
 - [Merge through a host gate](/quest/m0/host-land/README.md) - the host's fixer containers, and the frozen batched gate for the repositories not yet switched, with merge-commit landing as its one exception
 - [The land fixer resolves every mechanical failure](/quest/m0/land-fixer/README.md) - workflow pushes and the PR's own CI as context
 - [Reviewer usage limits](/quest/m0/review-quota.md) - `quest review` exits distinctly on a reviewer's quota and prints the reset time
+- [Review sessions keyed by upstream branch](/quest/m0/review-session-key.md) - a round from a local branch tracking the PR branch continues that branch's review session
 - [Kix decision UI](/quest/m0/kix/README.md) - planning and implementation through prompts and summaries, run on the maintainer's own runners
 - [Codex cloud off](/quest/m0/codex-cloud-off.md) - automatic Codex cloud reviews stop once no open PR relies on them
 - [Codex project-config skip](/quest/m0/codex-project-config.md) - once Codex keeps a supported way to skip a project's `.codex/` layer, reviews keep the user's config
diff --git a/quest/m0/review-session-key.md b/quest/m0/review-session-key.md
new file mode 100644
index 00000000..703f76f0
--- /dev/null
+++ b/quest/m0/review-session-key.md
@@ -0,0 +1,47 @@
+# [S] Review sessions keyed by upstream branch
+
+## Goal
+
+A `quest review` round run from a local branch that tracks the PR's branch,
+such as a temporary branch made because another worktree holds the PR branch,
+continues the PR branch's review session, so its round count and design
+checkpoint carry over. The iterate and review skills tell agents what to do
+when another worktree holds the PR branch.
+
+## Plan
+
+Decided 2026-11-04 (user, /quest-complete follow-up of
+[moq-dev/quest#474](https://github.com/moq-dev/quest/pull/474)). That day,
+iterate agents on [moq-dev/quest#464](https://github.com/moq-dev/quest/pull/464),
+[moq-dev/quest#469](https://github.com/moq-dev/quest/pull/469),
+[moq-dev/quest#472](https://github.com/moq-dev/quest/pull/472), and
+[moq-dev/quest#474](https://github.com/moq-dev/quest/pull/474) found the PR
+branch checked out in another worktree. Each reviewed from a temporary
+local branch, so `quest review`, which keys sessions by the checked-out branch
+name (`sessionBranch` and `sessionFile` in `src/cli/review.ts`), started a new
+session that resets the round-4 design checkpoint, or later resumed the wrong
+one.
+
+- What should the quest's goal be?
+  - ✅ Key by upstream branch: without `--branch`, the session branch is the
+    local branch's upstream branch name when it has one, else its own name.
+    The skills say to check out a PR branch another worktree holds detached
+    and pass `--branch <pr-branch>`.
+  - Rejected: skill guidance only; keying sessions by PR number through the forge
+- Which milestone?
+  - ✅ m0
+  - Rejected: m1; m2
+
+Implementation guidance:
+
+- The upstream's branch name, without the remote, is what matters, so a
+  fork's tracking branch keys the same session as the PR's head branch.
+- An upstream that is the review's base branch (`--base`, or origin's default
+  branch) names no PR, so such a branch keys by its own name. Branches made
+  with `git switch -c <name> origin/main` track the base by default (23 local
+  branches in the maintainer's clone did on 2026-11-04, this plan's own
+  included); keying them by upstream would merge every one into a single
+  `main` session. `review` names the session before it resolves the default
+  base today, so the base's branch has to be known first.
+- Nothing is in production, so sessions already stored under temporary names
+  need no migration.