M4: close out Save Points, with browser verification

Closes M4. The review's three findings are fixed, the durability rule the
specification always implied is now enforced, and M3's and M4's browser
behaviour has been verified in a real browser for the first time.

B-1 -- the Save Point list was an N+1 that loaded whole Action rows,
narration included, to answer "does a row exist here". It is now one bulk
two-column coordinate query plus one lineage: 53 SELECTs for 25 Save Points
became 5, and the count no longer grows with the list. The clause is an OR
of exact (branch, depth) pairs rather than two IN lists, because the cross
product would report a Save Point resolved on the strength of another one's
depth existing on this one's branch. A test builds exactly that trap.

B-2 -- reclassified during closeout from "missing warning" to a behaviour
defect, and fixed as one. STORY-BRANCH-SEMANTICS §19 says a named checkpoint
remains until explicitly deleted, and §28 already required future cleanup to
retain checkpoint-referenced paths; a cascade that silently removed Save
Points with a branch violated both, and a warning would only have documented
the violation. A branch a Save Point names can no longer be deleted. The
request is refused with the offending Save Points named, the user deletes
them explicitly -- which deletes no story -- and the branch then goes. The
scope is the subtree, because deleting a branch takes its descendants. Both
delete controls disable and explain. Recorded as a new §19.1; models.py,
TECHNICAL-DESIGN §8.8 and DATA-MODEL §8 had all recorded the cascade as the
rule and now record the refusal.

An earlier pass in this same closeout had kept the cascade and added a
warning. That was the wrong fix and its tests were replaced rather than left
standing, since they pinned the defect.

B-3 -- the D11/L03 automation never left one process, so it could not
distinguish durable state from a live Python object. It now spawns real
server processes, kills the first, and reads the campaign back with the
second.

C-5 -- creating a Save Point takes the campaign's turn lock. "Save where I
am" has to name one committed position, and the head is what a turn in
flight is about to move. Rename and Delete deliberately do not take it.

The architecture is untouched: a Save Point is still name + note +
(branch, depth), and restore is still coordinate -> head.move_to_node ->
head.move_to -> attempts.restore_state. No second restore path, no state
copied into a checkpoint, no fork on restore.

Browser verification -- the first in this project, and it covers both
milestones. Firefox 154.0.1 through geckodriver over the W3C WebDriver
protocol, driving the rendered DOM: 47/47 checks, twice, on independent
databases, no console errors. M3's Undo/Redo enable states, transcript
movement, Retry and the take pager, divergence retiring Redo; M4's whole
Save Point lifecycle, both confirmations, and the new branch-delete refusal
including its recovery. No dependency was added: the WebDriver client is
stdlib HTTP.

No application defect was found by the browser. Four failures occurred, all
in the harness -- a wrong SPA route, a wait comparing transcript length when
the empty-story placeholder is longer than the first turn, a fixture
deleting the branch it was reading, and a reload assertion that sampled
once instead of waiting. The last was checked against the app before being
called a harness bug.

Tests: 698 backend pass (was 680), 60 M4, 94 M3 history, 66 export/
migrations, 93 security/local-only. Frontend lint and build clean, Docker
build clean, loopback binding unchanged. No assertion weakened, no skip
added.

Planning: STORY-BRANCH-SEMANTICS §19.1 is the only behavioural change and it
strengthens §19. V1-ACCEPTANCE-TESTS records D11-D14, I04, L03 and the
E-series, keeping automated, live-runtime and browser evidence distinct, and
weakens no pass condition. DATA-MODEL records the coordinate with the retry
measurement that settles it. BROWSER-UX-SPEC rules for Moment over Turn.
BUILD-MILESTONES marks M4 COMPLETE, closes M3's browser condition, and lists
what M5 inherits. VERSION adds v2.6.

No new ADR: ADR 005 already decides that history is preserved rather than
overwritten, and §19.1 is that decision applied to checkpoint-referenced
history.

M4 is closed. M5 may now be briefed; it has not been started.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PWU4gTfLYY6Qq9U7aa9Qw2

DEVELOPMENT.md

@@ -209,7 +209,7 @@ visible from within.
 ## Tests
 
 ```bash
-cd backend && .venv/bin/python -m pytest tests/ -q     # 680 tests
+cd backend && .venv/bin/python -m pytest tests/ -q     # 698 tests
 cd frontend && npm run lint && npm run build
 ```
 
@@ -226,8 +226,15 @@ is lost from the union, or if a new HTTP client is added without the shared
 verification context.
 
 M4 added `test_save_points.py`, which fails if restoring a Save Point starts
-deleting history, stops going through the active head, forks on its own, or lets
+deleting history, stops going through the active head, forks on its own, lets a
-a Save Point on one campaign be restored through another.
+Save Point on one campaign be restored through another, or lets deleting a branch
+take a Save Point with it. It also fails if listing Save Points goes back to one
+query per Save Point, or starts fetching narration to render the list.
+
+`test_process_restart.py` is the durability guard: it starts the application as a
+real subprocess, kills it, and starts a second one against the same database. A
+Save Point that survived only because a Python object was still alive would pass
+an in-process test and fail a user's restart.
 
 M2 added two more. `test_endpoint_policy.py` fails if the set of reachable
 addresses widens, or if either place the rule is applied stops applying it —

README.md

@@ -74,8 +74,10 @@ that isn't the live one starts a new branch.
   is what starts a new line while the old one is kept. A Save Point is a name for
   a position and holds no copy of the story, so restoring it is the same
   movement Undo makes (`backend/app/routers/adventures/checkpoints.py`,
-  `backend/app/head.py`). They last until you delete them, and deleting one
+  `backend/app/head.py`). They last until *you* delete them: deleting one deletes
-  deletes no story.
+  no story, and deleting a branch a Save Point is kept on is refused until you
+  remove the Save Point yourself, so nothing takes a named moment away behind
+  your back.
 - **Import and export.** AI Dungeon-compatible scenario format; JSON for everything else. An adventure exports as `ai-dnd-adventure-v2`, which carries the whole tree:
   every branch, every take, the fork points, which branches the story has left behind, the Save
   Points and the position it is being read at — all of them chosen rather than computed, which is
@@ -98,8 +100,9 @@ that isn't the live one starts a new branch.
 
 None yet. The inherited screenshots showed upstream's UI — a Scripts tab, Log in and Sign up,
 a guest banner, scripting demo scenarios — none of which this fork has since M2, so they were
-removed rather than left standing as a picture of a product that no longer exists. New ones
+removed rather than left standing as a picture of a product that no longer exists. The M4
-are taken when the browser smoke test M3 still owes is run.
+closeout drove the real application in a real browser, so the screens exist and work; taking
+presentable screenshots of them is a job for the UI pass in M8.
 
 ## Quick start
 
@@ -221,7 +224,7 @@ development, Vite proxies `/api` to FastAPI.
 
 ## Tests
 
-680 backend tests: unit tests plus full HTTP integration through the real turn engine, with
+698 backend tests: unit tests plus full HTTP integration through the real turn engine, with
 the model provider mocked. They run with no route to the Internet, which is a requirement
 rather than a convenience — an offline claim proved on a machine that has been online once
 proves nothing.

backend/app/models.py

@@ -265,11 +265,18 @@ class Checkpoint(Base):
     resting in a shared prefix, and the node's own branch is the one that still
     names the position after the reader has moved elsewhere.
 
-    Deleting a branch deletes its Save Points, by the same cascade that takes
+    **Nothing removes a Save Point but the user.** They are not cleaned up for
-    its memories: the story the pointer names is gone with it. Nothing else
+    going stale, for being behind the head, or for pointing into a future the
-    removes one. They are not cleaned up for going stale, for being behind the
+    story has left (`STORY-BRANCH-SEMANTICS.md` §19).
-    head, or for pointing into a future the story has left
+
-    (`STORY-BRANCH-SEMANTICS.md` §19).
+    That includes deleting a branch. `branch_id` carries `ON DELETE CASCADE` as
+    referential integrity — a Save Point must never point at a branch that is
+    gone — but the branch endpoint refuses to delete a branch any Save Point
+    names, so the cascade does not fire through the application
+    (`routers/adventures/branches.py`, `STORY-BRANCH-SEMANTICS.md` §19.1). The
+    user deletes the Save Point first, which deletes no story, and then the
+    branch. Deleting the whole campaign does cascade, and should: that is what
+    the user asked for.
     """
 
     __tablename__ = "checkpoints"

backend/app/routers/adventures/branches.py

@@ -56,6 +56,19 @@ def list_branches(
         .group_by(models.Action.branch_id)
         .all()
     }
+    # M4 closeout: how many Save Points name a position on each line. Deleting a
+    # branch deletes them along with its story, and the panel has to be able to
+    # say so before the button is pressed (review §R B-2). One grouped query for
+    # the whole tree, like the one above it — never one per branch.
+    save_points = {
+        branch_id: count
+        for branch_id, count in db.query(
+            models.Checkpoint.branch_id, func.count(models.Checkpoint.id)
+        )
+        .filter(models.Checkpoint.adventure_id == adventure.id)
+        .group_by(models.Checkpoint.branch_id)
+        .all()
+    }
     out = []
     for branch in branches:
         count, tip = owned.get(branch.id, (0, None))
@@ -70,6 +83,7 @@ def list_branches(
                 branch.fork_depth if branch.fork_depth is not None else tree.NO_DEPTH
             ),
             own_actions=count,
+            save_points=save_points.get(branch.id, 0),
             is_head=(branch.id == adventure.head_branch_id),
             name=branch.name,
             created_at=branch.created_at,
@@ -151,15 +165,31 @@ def delete_branch(
     heavily retried adventure from growing without bound. That is why it ships
     with the view that first lets anyone create a fork rather than after it.
 
-    Two kinds of branch cannot be deleted. The root cannot, because it holds the
+    Three kinds of branch cannot be deleted. The root cannot, because it holds
-    turns every other branch borrows, so deleting it deletes the whole story. The
+    the turns every other branch borrows, so deleting it deletes the whole story.
-    branch currently being read cannot, and neither can any branch it was forked
+    The branch currently being read cannot, and neither can any branch it was
-    from, because the cascade would remove the head under the player and leave
+    forked from, because the cascade would remove the head under the player and
-    `head_branch_id` dangling. Switch branches first.
+    leave `head_branch_id` dangling. Switch branches first.
+
+    The third is M4's: **a branch a Save Point names cannot be deleted while that
+    Save Point exists.** `STORY-BRANCH-SEMANTICS.md` §19 says a named checkpoint
+    remains until explicitly deleted, and §28 says a future cleanup feature must
+    retain paths referenced by checkpoints. A cascade that removed Save Points
+    along with a branch would break both, and would break them silently: the
+    story the user asked to delete is the visible thing, and the named moments
+    would go without ever being named in the request. So the deletion is refused,
+    the Save Points are listed, and the user decides — delete the Save Point
+    first, then the branch. Deleting a Save Point still deletes no story (§25),
+    so the recovery costs nothing but a click.
+
+    The check covers the whole doomed subtree, not just this branch, because
+    deleting a branch takes everything forked from it.
 
     Nodes and memories are deleted by `ON DELETE CASCADE`, and descendants by the
     cascade on `branches.parent_branch_id`, so the delete is a single statement
-    however deep the subtree is.
+    however deep the subtree is. `checkpoints.branch_id` also carries a cascade,
+    as referential integrity — a Save Point must never point at a branch that is
+    gone — but the guard above means it does not fire through this endpoint.
     """
     branch = get_branch_or_404(adventure, branch_id, db)
     if branch.parent_branch_id is None:
@@ -178,6 +208,11 @@ def delete_branch(
             400, "You are reading this branch, or one forked from it. Switch to "
                  "another branch first.",
         )
+    # Refused before the lock is taken: this is a decision about the request, not
+    # a race with a turn.
+    protecting = _save_points_protecting(db, adventure, branch)
+    if protecting:
+        raise HTTPException(409, _protected_message(protecting))
     turns.acquire_turn_lock(adventure_id)
     try:
         # Collect the subtree before the delete, because afterwards there is no
@@ -199,6 +234,55 @@ def delete_branch(
     finally:
         turns._active_turns.discard(adventure_id)
 
+# How many Save Point names to spell out before the message starts summarising.
+# Enough to be actionable, few enough to stay a sentence.
+NAMED_IN_REFUSAL = 3
+
+
+def _save_points_protecting(
+    db: Session, adventure: models.Adventure, branch: models.Branch
+) -> list[models.Checkpoint]:
+    """Returns the Save Points that deleting `branch` would destroy.
+
+    The whole subtree, because deleting a branch takes everything forked from
+    it, and a check that looked only at this branch would let a Save Point on a
+    child be deleted without a word.
+    """
+    doomed = _branch_subtree(db, adventure, branch)
+    return (
+        db.query(models.Checkpoint)
+        .filter(
+            models.Checkpoint.adventure_id == adventure.id,
+            models.Checkpoint.branch_id.in_(doomed),
+        )
+        .order_by(models.Checkpoint.created_at, models.Checkpoint.id)
+        .all()
+    )
+
+
+def _protected_message(protecting: list[models.Checkpoint]) -> str:
+    """Says which Save Points stand in the way, and what to do about it.
+
+    Named rather than counted, because "2 Save Points" leaves the user hunting
+    for which ones. A long list is truncated so the message stays readable; the
+    Save Points panel shows the rest.
+    """
+    names = [f"“{c.name}”" for c in protecting[:NAMED_IN_REFUSAL]]
+    listed = ", ".join(names)
+    extra = len(protecting) - len(names)
+    if extra > 0:
+        listed += f" and {extra} more"
+    subject = "a Save Point" if len(protecting) == 1 else "Save Points"
+    return (
+        f"This branch, or a branch forked from it, is where {subject} "
+        f"{listed} {'is' if len(protecting) == 1 else 'are'} saved. Delete "
+        f"{'that Save Point' if len(protecting) == 1 else 'those Save Points'} "
+        f"first if you no longer need "
+        f"{'it' if len(protecting) == 1 else 'them'}, then delete the branch. "
+        f"Deleting a Save Point does not delete any story."
+    )
+
+
 def _branch_subtree(
     db: Session, adventure: models.Adventure, root: models.Branch
 ) -> set[int]:

backend/app/routers/adventures/checkpoints.py

@@ -23,7 +23,10 @@ The user-facing word is "Save Point" and the internal one is `checkpoint`
 Save Point.
 """
 
+from dataclasses import dataclass
+
 from fastapi import Depends, HTTPException
+from sqlalchemy import and_, or_
 from sqlalchemy.orm import Session
 
 from ... import head, models, schemas
@@ -64,18 +67,92 @@ def _node_at(
     )
 
 
+@dataclass(frozen=True)
+class _Coordinate:
+    """The shape `lineage.Path` reads, without loading a story row.
+
+    `Path.contains` asks three things of a node: its branch, its depth, and
+    whether it is live. A coordinate already known to resolve has all three, so
+    the membership question can be put to the coordinate itself. That keeps the
+    single implementation of "is this on the path being read" in `lineage`,
+    where M3 put it, while costing no query and no prose.
+    """
+
+    branch_id: int
+    depth: int
+    live: bool = True
+
+
+def _live_coordinates(
+    db: Session, adventure: models.Adventure, checkpoints: list[models.Checkpoint]
+) -> set[tuple[int, int]]:
+    """Returns which of these Save Points' coordinates still name a live turn.
+
+    One query for the whole list, selecting two integer columns.
+
+    This replaces a resolution per Save Point (M4 review §R B-1), which cost one
+    query each and loaded whole `Action` entities — narration included — to
+    answer a question that is only ever "does a row exist here". `paging.py`
+    states the rule this now follows: a bulk read names the columns it needs, so
+    a new column costs nothing until someone adds it to the list.
+
+    The clause is an OR of exact `(branch, depth)` pairs rather than
+    `branch IN (…) AND depth IN (…)`, which would match the cross product and
+    report a Save Point as resolved because *some other* Save Point's depth
+    exists on *this* one's branch.
+    """
+    coordinates = {(c.branch_id, c.depth) for c in checkpoints}
+    if not coordinates:
+        return set()
+    rows = (
+        db.query(models.Action.branch_id, models.Action.depth)
+        .filter(
+            models.Action.adventure_id == adventure.id,
+            models.Action.live.is_(True),
+            or_(*[
+                and_(models.Action.branch_id == branch, models.Action.depth == depth)
+                for branch, depth in coordinates
+            ]),
+        )
+        .all()
+    )
+    return {(branch, depth) for branch, depth in rows}
+
+
+def _render_all(
+    db: Session, adventure: models.Adventure, checkpoints: list[models.Checkpoint]
+) -> list[schemas.CheckpointOut]:
+    """Reads Save Points out with the three facts the panel needs about them.
+
+    Bounded work whatever the length of the list: one query for the coordinates
+    and one lineage for the campaign, both computed before the loop. Rendering
+    one Save Point and rendering fifty differ in Python, not in round trips.
+    """
+    live = _live_coordinates(db, adventure, checkpoints)
+    # The path is a property of the campaign, not of any Save Point, so it is
+    # read once. Reading it per row was the other half of the N+1.
+    path = lineage.path_of(db, adventure).uncapped()
+    out = []
+    for checkpoint in checkpoints:
+        coordinate = (checkpoint.branch_id, checkpoint.depth)
+        resolved = coordinate in live
+        rendered = schemas.CheckpointOut.model_validate(checkpoint)
+        # The same `depth + 1` the branch list counts with, so a moment number
+        # means the same thing in both places.
+        rendered.turn = checkpoint.depth + 1
+        rendered.resolved = resolved
+        rendered.on_path = resolved and path.contains(
+            _Coordinate(checkpoint.branch_id, checkpoint.depth)
+        )
+        out.append(rendered)
+    return out
+
+
 def _rendered(
     db: Session, adventure: models.Adventure, checkpoint: models.Checkpoint
 ) -> schemas.CheckpointOut:
-    """Reads one Save Point out with the three facts the panel needs about it."""
+    """Reads one Save Point out, through the same path the list uses."""
-    node = _node_at(db, adventure, checkpoint.branch_id, checkpoint.depth)
+    return _render_all(db, adventure, [checkpoint])[0]
-    out = schemas.CheckpointOut.model_validate(checkpoint)
-    # The same `depth + 1` the branch list counts with, so "turn 42" means the
-    # same thing in both places.
-    out.turn = checkpoint.depth + 1
-    out.resolved = node is not None
-    out.on_path = node is not None and lineage.path_of(db, adventure).uncapped().contains(node)
-    return out
 
 
 def _get_or_404(
@@ -123,7 +200,7 @@ def list_checkpoints(
         .order_by(models.Checkpoint.created_at.desc(), models.Checkpoint.id.desc())
         .all()
     )
-    return [_rendered(db, adventure, row) for row in rows]
+    return _render_all(db, adventure, rows)
 
 
 @router.post(
@@ -132,6 +209,7 @@ def list_checkpoints(
     status_code=201,
 )
 def create_checkpoint(
+    adventure_id: int,
     payload: schemas.CheckpointCreate,
     db: Session = Depends(get_db),
     adventure: models.Adventure = Depends(current_adventure),
@@ -148,22 +226,37 @@ def create_checkpoint(
     head resting in a shared prefix sits on an ancestor's node, and the
     ancestor is the branch that still names that position after the reader has
     forked away from it.
+
+    **Held under the campaign's turn lock** (M4 closeout, review §S C-5). "Save
+    where I am" has to name one committed position, and the head is exactly what
+    a turn in flight is about to move. Without the lock this endpoint could read
+    `head_depth` while a turn was mid-commit and store a coordinate for a
+    position the story had already left — a Save Point silently naming the wrong
+    moment, which no later operation could detect. It is the same lock Undo,
+    Redo and Restore take, for the same reason, and not a new mechanism.
+
+    Rename and Delete deliberately do **not** take it: neither reads nor moves a
+    story position, so there is nothing for a turn in flight to race them over.
     """
     name = _clean_name(payload.name)
-    node = head.node_at(db, adventure, adventure.head_depth)
+    turns.acquire_turn_lock(adventure_id)
-    if node is None:
+    try:
-        raise HTTPException(400, "There is no turn here to save yet.")
+        node = head.node_at(db, adventure, adventure.head_depth)
-    checkpoint = models.Checkpoint(
+        if node is None:
-        adventure_id=adventure.id,
+            raise HTTPException(400, "There is no turn here to save yet.")
-        name=name,
+        checkpoint = models.Checkpoint(
-        note=payload.note or "",
+            adventure_id=adventure.id,
-        branch_id=node.branch_id,
+            name=name,
-        depth=node.depth,
+            note=payload.note or "",
-    )
+            branch_id=node.branch_id,
-    db.add(checkpoint)
+            depth=node.depth,
-    db.commit()
+        )
-    db.refresh(checkpoint)
+        db.add(checkpoint)
-    return _rendered(db, adventure, checkpoint)
+        db.commit()
+        db.refresh(checkpoint)
+        return _rendered(db, adventure, checkpoint)
+    finally:
+        turns._active_turns.discard(adventure_id)
 
 
 @router.patch(

backend/app/schemas.py

@@ -259,6 +259,10 @@ class BranchOut(ORMModel):
     fork_depth: int | None = None
     depth: int
     own_actions: int = 0
+    # M4: how many Save Points name a position on this line. Deleting the branch
+    # deletes them with its story, so the panel warns with a number rather than
+    # a vague caution. Zero for a line nobody has bookmarked, which is most.
+    save_points: int = 0
     is_head: bool = False
     # NULL for a branch nobody has named. The client labels those from the fork
     # depth rather than the server inventing a name. See the column comment.

backend/tests/_restart_server.py

@@ -0,0 +1,63 @@
+"""The storyteller, run as a real OS process for `test_process_restart.py`.
+
+Not a test module, and named so pytest does not collect it: it is the program
+the test starts, twice, against one database file.
+
+The model is replaced with a deterministic fake before the app is imported, so
+the process needs no Ollama, no network and no configuration. Everything else —
+the engine, the migrations, the routers, the session lifecycle — is the real
+application, which is the whole point of spawning a process at all.
+
+    python _restart_server.py <db_path> <port>
+"""
+import itertools
+import os
+import sys
+from pathlib import Path
+
+HERE = Path(__file__).resolve().parent
+sys.path.insert(0, str(HERE.parent))   # backend/, so `app` imports
+sys.path.insert(0, str(HERE))          # tests/, so `fakes` imports
+
+db_path, port = sys.argv[1], int(sys.argv[2])
+os.environ["AIDND_DB_PATH"] = db_path
+# A developer's shell may point these at Postgres, and `app.database` prefers
+# either over the SQLite path. The suite's conftest clears them for the same
+# reason; a spawned process does not inherit that, so clear them here too.
+os.environ.pop("AIDND_DATABASE_URL", None)
+os.environ.pop("DATABASE_URL", None)
+
+from fakes import GOLD_PER_TURN, gold_reply  # noqa: E402
+
+_turn = itertools.count(1)
+
+
+class DeterministicProvider:
+    """Banks one turn's worth of gold per reply, numbered so text is checkable.
+
+    The same instrumentation `test_head_cursor.py` and `test_save_points.py`
+    use, for the same reason: it makes "the state at this position" a number the
+    test can assert rather than a paragraph it has to interpret. M5 replaces the
+    machinery underneath; what this measures is where the story is being read.
+    """
+
+    last_usage = None
+
+    def __init__(self, *a, **k):
+        pass
+
+    async def generate(self, parts, *, temperature, max_tokens):
+        yield ("text", gold_reply(f"Beat {next(_turn)}."))
+
+
+from app.routers.adventures import turns  # noqa: E402
+
+turns.OpenAICompatibleProvider = DeterministicProvider
+
+from app.main import app  # noqa: E402
+
+if __name__ == "__main__":
+    import uvicorn
+
+    # Loopback only, as every supported start path does.
+    uvicorn.run(app, host="127.0.0.1", port=port, log_level="warning")

backend/tests/test_process_restart.py

@@ -0,0 +1,325 @@
+"""D11 and L03 across a genuine OS process boundary.
+
+The rest of the suite runs the app in-process through `TestClient`, which is the
+right tool for almost everything and the wrong one for exactly one claim:
+*durability*. A Save Point that survived only because a Python object was still
+alive would pass a same-process test and fail a user's restart. `TestClient`
+cannot tell those apart, so the M4 review recorded the shipped D11/L03 tests as
+weaker than the acceptance items they were named for (§R B-3).
+
+This module closes that. It starts the real application as a **subprocess**,
+plays a story over HTTP, kills the process, starts a **second** process against
+the same database file, and only then asks whether the Save Point is still
+there. Everything crossing the boundary crosses it as bytes on disk.
+
+Deterministic and local: the spawned server replaces the model with a scripted
+provider (`_restart_server.py`), so there is no Ollama, no network and no
+sleep-and-hope — readiness is probed, not waited for.
+
+    python -m pytest tests/test_process_restart.py -v
+"""
+import json
+import os
+import socket
+import subprocess
+import sys
+import tempfile
+import time
+import urllib.error
+import urllib.request
+from pathlib import Path
+
+import pytest
+
+from fakes import GOLD_PER_TURN
+
+HERE = Path(__file__).resolve().parent
+SERVER = HERE / "_restart_server.py"
+
+# How long a spawned server may take to answer before the test gives up. The
+# process imports the app and runs migrations on a fresh file, which is well
+# under a second on this project; the ceiling is for a loaded machine.
+STARTUP_TIMEOUT = 60.0
+
+
+def _free_port() -> int:
+    """Returns a port nothing is listening on.
+
+    Bind, read, release. There is a race between releasing and the server
+    claiming it, which is why the caller probes for readiness rather than
+    assuming success — a lost race shows up as a startup timeout, not as a
+    silent pass.
+    """
+    with socket.socket() as s:
+        s.bind(("127.0.0.1", 0))
+        return s.getsockname()[1]
+
+
+class Server:
+    """One storyteller process, and the HTTP calls the test makes against it."""
+
+    def __init__(self, db_path: str, port: int):
+        self.port = port
+        self.proc = subprocess.Popen(
+            [sys.executable, str(SERVER), db_path, str(port)],
+            stdout=subprocess.PIPE,
+            stderr=subprocess.STDOUT,
+            # Never inherit the parent's database redirection; the child is told
+            # which file to open on its command line.
+            env={**os.environ, "AIDND_DB_PATH": db_path},
+        )
+
+    # ------------------------------------------------------------ lifecycle
+
+    def wait_until_ready(self) -> None:
+        deadline = time.monotonic() + STARTUP_TIMEOUT
+        while time.monotonic() < deadline:
+            if self.proc.poll() is not None:
+                raise AssertionError(
+                    f"server exited early ({self.proc.returncode}):\n{self._output()}"
+                )
+            try:
+                self.call("GET", "/settings")
+                return
+            except (urllib.error.URLError, ConnectionError, OSError):
+                time.sleep(0.05)
+        raise AssertionError(f"server never became ready:\n{self._output()}")
+
+    def stop(self) -> None:
+        """Ends the process, and does not return until it is actually gone."""
+        if self.proc.poll() is None:
+            self.proc.terminate()
+            try:
+                self.proc.wait(timeout=15)
+            except subprocess.TimeoutExpired:
+                self.proc.kill()
+                self.proc.wait(timeout=15)
+        if self.proc.stdout is not None:
+            self.proc.stdout.close()
+
+    def _output(self) -> str:
+        if self.proc.stdout is None:
+            return "(no output captured)"
+        try:
+            return self.proc.stdout.read().decode(errors="replace")[-2000:]
+        except Exception:
+            return "(output unreadable)"
+
+    def is_listening(self) -> bool:
+        try:
+            self.call("GET", "/settings")
+            return True
+        except Exception:
+            return False
+
+    # ---------------------------------------------------------------- HTTP
+
+    def call(self, method: str, path: str, payload=None, expect: int | None = None):
+        data = json.dumps(payload).encode() if payload is not None else None
+        request = urllib.request.Request(
+            f"http://127.0.0.1:{self.port}/api{path}",
+            data=data,
+            method=method,
+            headers={"Content-Type": "application/json"},
+        )
+        try:
+            with urllib.request.urlopen(request, timeout=60) as response:
+                body, status = response.read(), response.status
+        except urllib.error.HTTPError as exc:          # a real answer, not a failure
+            body, status = exc.read(), exc.code
+        if expect is not None and status != expect:
+            raise AssertionError(f"{method} {path} -> {status}: {body[:400]!r}")
+        return json.loads(body) if body and status != 204 else None
+
+    def play(self, adventure_id: int, text: str) -> None:
+        """Plays one turn through the streaming endpoint, to completion."""
+        request = urllib.request.Request(
+            f"http://127.0.0.1:{self.port}/api/adventures/{adventure_id}/actions",
+            data=json.dumps({"type": "do", "text": text}).encode(),
+            method="POST",
+            headers={"Content-Type": "application/json"},
+        )
+        with urllib.request.urlopen(request, timeout=120) as response:
+            response.read()
+
+    # ------------------------------------------------------------- reading
+
+    def transcript(self, adventure_id: int) -> list[str]:
+        page = self.call("GET", f"/adventures/{adventure_id}", expect=200)
+        return [a["text"] for a in page["actions"]]
+
+    def gold(self, adventure_id: int) -> int:
+        state = self.call("GET", f"/adventures/{adventure_id}/world-state", expect=200)
+        return state["state"]["player"]["gold"]
+
+    def total_rows(self, adventure_id: int) -> int:
+        """Every row of the whole tree, head or no head.
+
+        `action_count` is scoped to the path being read, so it falls when the
+        head moves back — which is the feature, not a deletion. The export
+        carries the entire tree whatever the head is doing, so it is what
+        "nothing was deleted" has to be measured against.
+        """
+        bundle = self.call("GET", f"/adventures/{adventure_id}/export", expect=200)
+        return len(bundle["actions"])
+
+
+@pytest.fixture()
+def workspace():
+    """A database file, and whichever servers a test starts against it."""
+    directory = tempfile.mkdtemp(prefix="m4-restart-")
+    db_path = os.path.join(directory, "campaign.db")
+    started: list[Server] = []
+
+    def start() -> Server:
+        server = Server(db_path, _free_port())
+        started.append(server)
+        server.wait_until_ready()
+        return server
+
+    try:
+        yield start
+    finally:
+        # Every child dies even if the test failed part way through, and each
+        # stop() waits, so a later test cannot inherit a live listener.
+        for server in started:
+            server.stop()
+
+
+def _campaign_with_a_save_point(server: Server):
+    """Three turns, a Save Point on the third, then four more turns.
+
+    Returns everything the second process has to be able to reproduce.
+    """
+    scenario = server.call("POST", "/scenarios", {
+        "title": "Abbey",
+        "stat_schema": {"player": {"gold": {"initial": 0, "min": 0, "max": 9999}}},
+    }, expect=201)
+    adventure = server.call("POST", "/adventures", {
+        "title": "The abbey", "scenario_id": scenario["id"],
+    }, expect=201)
+    adventure_id = adventure["id"]
+
+    for n in range(3):
+        server.play(adventure_id, f"turn {n}")
+    at_save = {
+        "transcript": server.transcript(adventure_id),
+        "gold": server.gold(adventure_id),
+    }
+    save_point = server.call("POST", f"/adventures/{adventure_id}/checkpoints",
+                             {"name": "Before entering the abbey"}, expect=201)
+
+    for n in range(4):
+        server.play(adventure_id, f"later {n}")
+    at_tip = {
+        "transcript": server.transcript(adventure_id),
+        "gold": server.gold(adventure_id),
+        "rows": server.total_rows(adventure_id),
+    }
+    return adventure_id, save_point, at_save, at_tip
+
+
+def test_d11_l03_a_save_point_survives_a_real_process_restart(workspace):
+    """D11 and L03 together, across a boundary a same-process test cannot cross.
+
+    The first process writes the campaign and exits. The second process is a
+    different interpreter with an empty session, an empty identity map and no
+    memory of anything — everything it knows, it reads off the disk.
+    """
+    first = workspace()
+    adventure_id, save_point, at_save, at_tip = _campaign_with_a_save_point(first)
+
+    assert at_tip["gold"] == at_save["gold"] + 4 * GOLD_PER_TURN
+    assert len(at_tip["transcript"]) == len(at_save["transcript"]) + 8
+
+    # --- the boundary -----------------------------------------------------
+    first.stop()
+    assert first.proc.poll() is not None, "the first server did not actually exit"
+    assert not first.is_listening(), "the first server is still answering"
+
+    second = workspace()
+    assert second.proc.pid != first.proc.pid
+
+    # --- D11: the Save Point is still there -------------------------------
+    listed = second.call("GET", f"/adventures/{adventure_id}/checkpoints", expect=200)
+    assert [c["id"] for c in listed] == [save_point["id"]]
+    assert listed[0]["name"] == "Before entering the abbey"
+    assert (listed[0]["branch_id"], listed[0]["depth"]) == (
+        save_point["branch_id"], save_point["depth"]
+    )
+    assert listed[0]["resolved"] is True
+    # The story came back too, at the position the first process left it.
+    assert second.transcript(adventure_id) == at_tip["transcript"]
+
+    rows_before_restore = second.total_rows(adventure_id)
+    assert rows_before_restore == at_tip["rows"]
+
+    # --- L03: restoring reconstructs the historical position --------------
+    page = second.call(
+        "POST", f"/adventures/{adventure_id}/checkpoints/{save_point['id']}/restore",
+        expect=200,
+    )
+    assert [a["text"] for a in page["actions"]] == at_save["transcript"]
+    assert second.transcript(adventure_id) == at_save["transcript"]
+    assert second.gold(adventure_id) == at_save["gold"]
+
+    # --- restore deleted nothing, and Redo is still available -------------
+    assert second.total_rows(adventure_id) == rows_before_restore
+    assert page["can_redo"] is True
+    assert second.call("GET", f"/adventures/{adventure_id}", expect=200)["can_redo"] is True
+
+
+def test_the_retained_continuation_is_reachable_after_a_restart(workspace):
+    """The later history is not merely present in the database after a restart —
+    it is still the continuation the story tells, walkable by ordinary Redo."""
+    first = workspace()
+    adventure_id, save_point, at_save, at_tip = _campaign_with_a_save_point(first)
+    first.stop()
+
+    second = workspace()
+    second.call("POST",
+                f"/adventures/{adventure_id}/checkpoints/{save_point['id']}/restore",
+                expect=200)
+    assert second.transcript(adventure_id) == at_save["transcript"]
+
+    steps = 0
+    while second.call("GET", f"/adventures/{adventure_id}", expect=200)["can_redo"]:
+        second.call("POST", f"/adventures/{adventure_id}/redo", expect=200)
+        steps += 1
+        assert steps <= 10, "Redo never stopped"
+
+    assert steps == 4
+    assert second.transcript(adventure_id) == at_tip["transcript"]
+    assert second.gold(adventure_id) == at_tip["gold"]
+    assert second.total_rows(adventure_id) == at_tip["rows"]
+
+
+def test_a_divergent_write_after_a_restart_keeps_the_displaced_future(workspace):
+    """D13's second half, across the boundary: the fork still happens on the
+    write rather than on the restore, and the displaced turns keep their rows."""
+    first = workspace()
+    adventure_id, save_point, at_save, at_tip = _campaign_with_a_save_point(first)
+    first.stop()
+
+    second = workspace()
+    second.call("POST",
+                f"/adventures/{adventure_id}/checkpoints/{save_point['id']}/restore",
+                expect=200)
+    branches_before = second.call("GET", f"/adventures/{adventure_id}/branches",
+                                  expect=200)
+    assert len(branches_before) == 1, "restore must not fork"
+
+    second.play(adventure_id, "go around the back")
+
+    branches_after = second.call("GET", f"/adventures/{adventure_id}/branches",
+                                 expect=200)
+    assert len(branches_after) == 2, "the write should have forked"
+    # Nothing was deleted to achieve it: the tree grew by the new turn only.
+    assert second.total_rows(adventure_id) == at_tip["rows"] + 2
+    # Ordinary Redo no longer offers the displaced future.
+    assert second.call("GET", f"/adventures/{adventure_id}", expect=200)["can_redo"] is False
+    # And the Save Point still names the position it always did.
+    listed = second.call("GET", f"/adventures/{adventure_id}/checkpoints", expect=200)
+    assert (listed[0]["branch_id"], listed[0]["depth"]) == (
+        save_point["branch_id"], save_point["depth"]
+    )

backend/tests/test_save_points.py

@@ -748,36 +748,169 @@ def test_a_save_point_whose_turn_is_gone_refuses_rather_than_approximating(clien
     assert _head(client.adv_id) == head_before
 
 
-def test_deleting_a_branch_takes_its_save_points_with_it(client):
+def test_a_branch_a_save_point_names_cannot_be_deleted(client):
-    """Referential integrity, not cleanup. Nothing removes a Save Point for
+    """`STORY-BRANCH-SEMANTICS.md` §19: a named Save Point remains until it is
-    going stale; this one goes because the story it named went."""
+    explicitly deleted — and §28 says even a future cleanup feature must retain
+    paths a checkpoint references.
+
+    So the branch delete is refused rather than taking the Save Point with it.
+    The alternative, a silent cascade, breaks §19 in the way that matters least
+    visibly: the story is the thing the user asked to delete, and the named
+    moments would go without ever being mentioned.
+    """
     _turns(client, 2)
     _undo(client)
     _undo(client)
     ScriptedProvider.replies = ["A new road.\n```state\n{\"player.gold\": 1}\n```"]
     _play(client, "the other way")
     forked = _save(client, "On the new line")
+    doomed_branch = forked["branch_id"]
 
-    # Read somewhere the doomed branch is not load-bearing, then delete it.
+    root_id = _root_branch(client.adv_id)
+    assert doomed_branch != root_id
+    client.post(f"/api/adventures/{client.adv_id}/branches/{root_id}/switch")
+
+    r = client.delete(f"/api/adventures/{client.adv_id}/branches/{doomed_branch}")
+    assert r.status_code == 409, r.text
+    detail = r.json()["detail"]
+    # The message names the Save Point, so the user does not have to hunt.
+    assert "On the new line" in detail
+    assert "does not delete any story" in detail
+
+    # Nothing happened: the Save Point, the branch and the story all remain.
+    assert [c["id"] for c in _list(client)] == [forked["id"]]
+    assert _branch_count(client.adv_id) == 2
+    assert any(a.branch_id == doomed_branch for a in _rows(client.adv_id))
+
+
+def test_deleting_the_save_point_then_lets_the_branch_go(client):
+    """The refusal has to be recoverable, or it is just a wall. Deleting the
+    Save Point deletes no story (§25), so the cost of the recovery is a click."""
+    _turns(client, 2)
+    _undo(client)
+    _undo(client)
+    ScriptedProvider.replies = ["A new road.\n```state\n{\"player.gold\": 1}\n```"]
+    _play(client, "the other way")
+    forked = _save(client, "On the new line")
+    doomed_branch = forked["branch_id"]
+    root_id = _root_branch(client.adv_id)
+    client.post(f"/api/adventures/{client.adv_id}/branches/{root_id}/switch")
+
+    assert client.delete(
+        f"/api/adventures/{client.adv_id}/branches/{doomed_branch}"
+    ).status_code == 409
+    rows_with_story = len(_rows(client.adv_id))
+
+    # Delete the Save Point explicitly...
+    assert client.delete(
+        f"/api/adventures/{client.adv_id}/checkpoints/{forked['id']}"
+    ).status_code == 204
+    # ...which took no story with it...
+    assert len(_rows(client.adv_id)) == rows_with_story
+
+    # ...and now the branch can go.
+    r = client.delete(f"/api/adventures/{client.adv_id}/branches/{doomed_branch}")
+    assert r.status_code in (200, 204), r.text
+    assert _branch_count(client.adv_id) == 1
+
+
+def test_a_branch_no_save_point_names_still_deletes(client):
+    """The guard must not turn into a general refusal to delete branches."""
+    _turns(client, 2)
+    _undo(client)
+    _undo(client)
+    ScriptedProvider.replies = ["A new road.\n```state\n{\"player.gold\": 1}\n```"]
+    _play(client, "the other way")
+    forked_branch = _head(client.adv_id)[0]
+    root_id = _root_branch(client.adv_id)
+    client.post(f"/api/adventures/{client.adv_id}/branches/{root_id}/switch")
+
+    assert _list(client) == []
+    r = client.delete(f"/api/adventures/{client.adv_id}/branches/{forked_branch}")
+    assert r.status_code in (200, 204), r.text
+    assert _branch_count(client.adv_id) == 1
+
+
+def test_a_save_point_on_a_descendant_also_protects_the_branch(client):
+    """Deleting a branch takes everything forked from it, so the check has to
+    cover the subtree. A guard that looked only at the named branch would let a
+    Save Point on a child be deleted without a word — the exact failure the
+    guard exists to prevent, one level down."""
+    _turns(client, 4)
+    _undo(client)
+    _undo(client)
+    ScriptedProvider.replies = ["Second line.\n```state\n{\"player.gold\": 1}\n```"]
+    _play(client, "second line")
+    middle_branch = _head(client.adv_id)[0]
+    _undo(client)
+    ScriptedProvider.replies = ["Third line.\n```state\n{\"player.gold\": 1}\n```"]
+    _play(client, "third line")
+    deepest = _save(client, "Down on the deepest line")
+    assert deepest["branch_id"] != middle_branch
+
+    root_id = _root_branch(client.adv_id)
+    client.post(f"/api/adventures/{client.adv_id}/branches/{root_id}/switch")
+
+    # Deleting the *middle* branch would take the deepest one with it.
+    r = client.delete(f"/api/adventures/{client.adv_id}/branches/{middle_branch}")
+    assert r.status_code == 409, r.text
+    assert "Down on the deepest line" in r.json()["detail"]
+    assert [c["id"] for c in _list(client)] == [deepest["id"]]
+
+
+def test_a_save_point_elsewhere_does_not_block_an_unrelated_branch(client):
+    """The guard is scoped to the subtree being deleted, not to the campaign."""
+    _turns(client, 3)
+    elsewhere = _save(client, "Safe on the root")
+    _undo(client)
+    _undo(client)
+    ScriptedProvider.replies = ["A new road.\n```state\n{\"player.gold\": 1}\n```"]
+    _play(client, "the other way")
+    forked_branch = _head(client.adv_id)[0]
+    root_id = _root_branch(client.adv_id)
+    client.post(f"/api/adventures/{client.adv_id}/branches/{root_id}/switch")
+
+    assert elsewhere["branch_id"] == root_id
+    r = client.delete(f"/api/adventures/{client.adv_id}/branches/{forked_branch}")
+    assert r.status_code in (200, 204), r.text
+    # The unrelated Save Point is untouched and still restores.
+    assert [c["id"] for c in _list(client)] == [elsewhere["id"]]
+    assert _restore(client, elsewhere["id"]).status_code == 200
+
+
+def test_the_refusal_names_several_save_points_without_running_on(client):
+    """A long list is truncated so the message stays a sentence someone reads."""
+    _turns(client, 2)
+    _undo(client)
+    _undo(client)
+    ScriptedProvider.replies = ["A new road.\n```state\n{\"player.gold\": 1}\n```"]
+    _play(client, "the other way")
+    for n in range(5):
+        _save(client, f"Point {n}")
+    doomed_branch = _head(client.adv_id)[0]
+    root_id = _root_branch(client.adv_id)
+    client.post(f"/api/adventures/{client.adv_id}/branches/{root_id}/switch")
+
+    detail = client.delete(
+        f"/api/adventures/{client.adv_id}/branches/{doomed_branch}"
+    ).json()["detail"]
+    assert "Point 0" in detail and "2 more" in detail
+    assert "Point 4" not in detail
+    assert len(_list(client)) == 5
+
+
+def _root_branch(adv_id) -> int:
     db = SessionLocal()
     try:
-        adv = db.get(models.Adventure, client.adv_id)
+        return (
-        root = (
             db.query(models.Branch)
-            .filter_by(adventure_id=client.adv_id, parent_branch_id=None)
+            .filter_by(adventure_id=adv_id, parent_branch_id=None)
             .one()
+            .id
         )
-        doomed_id = forked["branch_id"]
-        assert doomed_id != root.id
     finally:
         db.close()
 
-    client.post(f"/api/adventures/{client.adv_id}/branches/{root.id}/switch")
-    r = client.delete(f"/api/adventures/{client.adv_id}/branches/{doomed_id}")
-    assert r.status_code in (200, 204), r.text
-
-    assert [c["id"] for c in _list(client)] == []
-
 
 # --------------------------------------------- E-series: lineage and memory
 
@@ -1129,3 +1262,328 @@ def test_an_m3_database_gains_the_save_point_table_and_keeps_its_story():
     migrations.bootstrap(m3)
     assert "checkpoints" in inspect(m3).get_table_names()
     m3.dispose()
+
+
+# ------------------------------------------- the cost of listing Save Points
+
+def _sql_during(work):
+    """Returns every SQL statement a block of work executed."""
+    from sqlalchemy import event
+    seen: list[str] = []
+
+    def record(conn, cursor, statement, params, context, executemany):
+        seen.append(statement)
+
+    event.listen(engine, "before_cursor_execute", record)
+    try:
+        work()
+    finally:
+        event.remove(engine, "before_cursor_execute", record)
+    return seen
+
+
+def test_listing_save_points_costs_a_bounded_number_of_queries(client):
+    """M4 review §R B-1. The list resolved each Save Point on its own, so the
+    query count grew with the list: 53 SELECTs for 25 Save Points, against 4 for
+    the comparable branch panel.
+
+    The assertion is on *growth*, not on an exact number, because a fixed budget
+    would be a number to edit rather than a rule to keep. Five times the Save
+    Points must not mean five times the queries.
+    """
+    _turns(client, 5)
+    for n in range(5):
+        _save(client, f"Save Point {n}")
+    few = _sql_during(lambda: _list(client))
+
+    _turns(client, 20)
+    for n in range(20):
+        _save(client, f"Later Save Point {n}")
+    assert len(_list(client)) == 25
+    many = _sql_during(lambda: _list(client))
+
+    # Five times the rows, and the query count does not move at all.
+    assert len(many) == len(few), (
+        f"listing 25 Save Points cost {len(many)} queries where 5 cost {len(few)}"
+    )
+    # And the whole thing is a handful, not a per-row walk.
+    assert len(many) <= 6, f"{len(many)} queries to list 25 Save Points"
+
+
+def test_listing_save_points_does_not_read_story_prose(client):
+    """The other half of B-1. Resolving a coordinate asks whether a row exists;
+    it never needs the narration in it, and `paging.py` states the rule this
+    follows — a bulk read names the columns it needs.
+
+    Enforced on the emitted SQL rather than on a byte count, because the failure
+    this guards against is a `SELECT` widening back to the whole entity, which a
+    small fixture would not make visible in bytes.
+    """
+    _turns(client, 3)
+    for n in range(3):
+        _save(client, f"Save Point {n}")
+
+    statements = _sql_during(lambda: _list(client))
+    action_reads = [q for q in statements if "FROM actions" in q]
+    assert action_reads, "the list must still check that coordinates resolve"
+    for query in action_reads:
+        selected = query.split("FROM actions")[0]
+        for column in ("actions.text", "actions.reasoning", "actions.world_delta"):
+            assert column not in selected, f"{column} fetched to render the list:\n{query}"
+
+
+def test_a_save_point_on_a_deleted_turn_is_still_reported_unresolved(client):
+    """The bulk resolution must not have quietly changed what `resolved` means.
+
+    This is the negative control for the B-1 rewrite: one query for many
+    coordinates is only correct if a coordinate with no live row still comes
+    back false.
+    """
+    _turns(client, 3)
+    alive = _save(client, "Still here")
+    _turns(client, 1)
+    doomed = _save(client, "About to vanish")
+
+    db = SessionLocal()
+    try:
+        for row in db.query(models.Action).filter_by(
+            adventure_id=client.adv_id,
+            branch_id=doomed["branch_id"],
+            depth=doomed["depth"],
+        ):
+            db.delete(row)
+        db.commit()
+    finally:
+        db.close()
+
+    by_id = {c["id"]: c for c in _list(client)}
+    assert by_id[doomed["id"]]["resolved"] is False
+    assert by_id[alive["id"]]["resolved"] is True
+    # One resolving and one not, in the same single query.
+
+
+def test_the_bulk_resolution_does_not_confuse_coordinates_across_branches(client):
+    """A coordinate is a pair, and the bulk query must match it as a pair.
+
+    Matching `branch IN (...) AND depth IN (...)` would take the cross product,
+    and a Save Point at a depth that exists on *another* branch would be
+    reported as resolved. This builds exactly that trap: two branches, and a
+    Save Point whose own coordinate is dead while the other branch has a live
+    row at the same depth.
+    """
+    _turns(client, 4)
+    _undo(client)
+    _undo(client)
+    ScriptedProvider.replies = ["A new road.\n```state\n{\"player.gold\": 1}\n```"]
+    _play(client, "the other way")          # forks; new branch has rows at 5,6
+    on_new_line = _save(client, "On the new line")
+
+    # A Save Point on the old line at the same depth, whose row we then remove.
+    db = SessionLocal()
+    try:
+        adventure = db.get(models.Adventure, client.adv_id)
+        old_branch = (
+            db.query(models.Branch)
+            .filter_by(adventure_id=client.adv_id, parent_branch_id=None)
+            .one()
+        )
+        stranded = models.Checkpoint(
+            adventure_id=client.adv_id,
+            name="Stranded on the old line",
+            branch_id=old_branch.id,
+            depth=on_new_line["depth"],
+        )
+        db.add(stranded)
+        db.flush()
+        stranded_id = stranded.id
+        # Remove the old line's row at that depth, so this coordinate is dead
+        # while the *other* branch still has a live row at the same depth.
+        for row in db.query(models.Action).filter_by(
+            adventure_id=client.adv_id,
+            branch_id=old_branch.id,
+            depth=on_new_line["depth"],
+        ):
+            db.delete(row)
+        db.commit()
+    finally:
+        db.close()
+
+    by_id = {c["id"]: c for c in _list(client)}
+    assert by_id[on_new_line["id"]]["resolved"] is True
+    assert by_id[stranded_id]["resolved"] is False, (
+        "a dead coordinate was reported resolved because another branch has a "
+        "live row at the same depth"
+    )
+
+
+# --------------------------------- deleting a branch, and saying so first
+
+def test_the_branch_list_reports_how_many_save_points_each_line_carries(client):
+    """The number the delete warning is built from (M4 review §R B-2).
+
+    Served as part of the branch list rather than from an endpoint of its own,
+    and as one grouped query rather than one per branch — the panel already
+    reads this list to draw itself.
+    """
+    _turns(client, 2)
+    _undo(client)
+    _undo(client)
+    ScriptedProvider.replies = ["A new road.\n```state\n{\"player.gold\": 1}\n```"]
+    _play(client, "the other way")
+    on_new = _save(client, "On the new line")
+
+    branches = client.get(f"/api/adventures/{client.adv_id}/branches").json()
+    by_id = {b["id"]: b for b in branches}
+    assert len(branches) == 2
+    assert by_id[on_new["branch_id"]]["save_points"] == 1
+    other = next(b for b in branches if b["id"] != on_new["branch_id"])
+    assert other["save_points"] == 0
+
+    _save(client, "A second one here")
+    branches = client.get(f"/api/adventures/{client.adv_id}/branches").json()
+    assert {b["id"]: b["save_points"] for b in branches}[on_new["branch_id"]] == 2
+
+
+def test_the_save_point_count_matches_what_blocks_the_deletion(client):
+    """The number the panel disables its Delete button on has to be the same
+    number the server refuses on, or the UI and the rule disagree."""
+    _turns(client, 4)
+    _undo(client)
+    _undo(client)
+    ScriptedProvider.replies = ["Second line.\n```state\n{\"player.gold\": 1}\n```"]
+    _play(client, "second line")
+    middle_branch = _head(client.adv_id)[0]
+    _undo(client)
+    ScriptedProvider.replies = ["Third line.\n```state\n{\"player.gold\": 1}\n```"]
+    _play(client, "third line")
+    _save(client, "Deep one")
+
+    branches = client.get(f"/api/adventures/{client.adv_id}/branches").json()
+    counts = {b["id"]: b["save_points"] for b in branches}
+    # The count is per branch; the client sums it over the subtree, and the
+    # server refuses on the same subtree.
+    assert sum(counts.values()) == 1
+    assert counts[middle_branch] == 0
+
+    root_id = _root_branch(client.adv_id)
+    client.post(f"/api/adventures/{client.adv_id}/branches/{root_id}/switch")
+    assert client.delete(
+        f"/api/adventures/{client.adv_id}/branches/{middle_branch}"
+    ).status_code == 409
+
+
+def test_both_branch_delete_surfaces_explain_the_save_point_rule():
+    """The rule must be visible in every view the deletion is reachable from.
+
+    A source-level assertion, because the project has no frontend test runner
+    (M8). `test_offline_assets.py` reads the frontend the same way, for the same
+    reason: the check is worth having now, and it is honest about what it is —
+    it proves the wiring is in the build, not that a user saw it. The browser
+    smoke test performed at closeout is what proves the rendering.
+
+    Two files, because the branch list and the tree overlay each render their
+    own delete control, and a rule that held in one of them would not be a rule.
+    """
+    from pathlib import Path
+
+    repo = Path(__file__).resolve().parents[2]
+    views = {
+        "the branch panel":
+            repo / "frontend/src/pages/Play/panels/BranchPanel.jsx",
+        "the tree overlay":
+            repo / "frontend/src/BranchMap.jsx",
+    }
+    for where, path in views.items():
+        source = path.read_text(encoding="utf-8")
+        assert "savePointsUnder" in source, (
+            f"{where} does not count the Save Points that protect a branch"
+        )
+        # The Delete control is disabled while Save Points protect the subtree,
+        # and says why rather than failing silently on the server.
+        assert "protecting > 0" in source, (
+            f"{where} does not disable Delete while Save Points protect the branch"
+        )
+        assert "deleting a Save Point deletes no story" in source, (
+            f"{where} does not tell the user how to proceed"
+        )
+        # The user-facing copy must not explain itself in schema terms.
+        for jargon in ("cascade", "foreign key", "foreign-key", "ON DELETE"):
+            assert jargon.lower() not in source.lower(), (
+                f"{where} uses implementation jargon in user-facing copy: {jargon}"
+            )
+
+
+def test_creating_a_save_point_takes_the_campaigns_turn_lock(client):
+    """M4 review §S C-5. "Save where I am" has to name one committed position,
+    and the head is exactly what a turn in flight is about to move.
+
+    Asserted by holding the lock and watching Create refuse, which is the same
+    contract Undo, Redo and Restore already answer with a 409. That proves it
+    participates in the serialization rather than merely being fast.
+    """
+    _turns(client, 2)
+    adventures.turns.acquire_turn_lock(client.adv_id)
+    try:
+        r = client.post(f"/api/adventures/{client.adv_id}/checkpoints",
+                        json={"name": "During a turn"})
+        assert r.status_code == 409, r.text
+        # The same refusal the other position-moving operations give.
+        assert _undo(client).status_code == 409
+        assert _restore(client, 1).status_code in (404, 409)
+    finally:
+        adventures.turns._active_turns.discard(client.adv_id)
+
+    # Nothing was written while the lock was held...
+    assert _list(client) == []
+    # ...and the endpoint works again once the turn is done.
+    assert _save(client, "After the turn")["name"] == "After the turn"
+
+
+def test_creating_a_save_point_releases_the_lock_even_when_it_refuses(client):
+    """A refused create must not leave the campaign wedged.
+
+    The empty-story refusal is raised from inside the locked section, so this is
+    the case that would strand the lock if the release were not in a `finally`.
+    """
+    db = SessionLocal()
+    try:
+        adv = db.get(models.Adventure, client.adv_id)
+        adv.head_depth = lineage.NO_DEPTH
+        db.commit()
+    finally:
+        db.close()
+
+    r = client.post(f"/api/adventures/{client.adv_id}/checkpoints",
+                    json={"name": "Nowhere"})
+    assert r.status_code == 400
+
+    # The lock is free: an ordinary turn still works.
+    assert client.adv_id not in adventures.turns._active_turns
+    _play(client, "carry on")
+
+
+def test_rename_and_delete_do_not_need_the_turn_lock(client):
+    """Deliberate, and worth pinning down: neither reads nor moves a story
+    position, so neither can race a turn in flight. Renaming a Save Point while
+    a turn generates is a label edit, and refusing it would be a worse product
+    for no safety gained."""
+    _turns(client, 2)
+    made = _save(client, "One")
+
+    adventures.turns.acquire_turn_lock(client.adv_id)
+    try:
+        renamed = client.patch(
+            f"/api/adventures/{client.adv_id}/checkpoints/{made['id']}",
+            json={"name": "Renamed mid-turn"},
+        )
+        assert renamed.status_code == 200, renamed.text
+        assert renamed.json()["name"] == "Renamed mid-turn"
+        # And the coordinate did not move while a turn was in flight.
+        assert (renamed.json()["branch_id"], renamed.json()["depth"]) == (
+            made["branch_id"], made["depth"]
+        )
+        assert client.delete(
+            f"/api/adventures/{client.adv_id}/checkpoints/{made['id']}"
+        ).status_code == 204
+    finally:
+        adventures.turns._active_turns.discard(client.adv_id)

frontend/src/BranchMap.jsx

@@ -1,6 +1,6 @@
 import { useEffect, useLayoutEffect, useMemo, useRef, useState } from 'react'
 import { createPortal } from 'react-dom'
-import { CORNER, PAD, ROW_H, branchLabel, headLineage, layoutTree, momentTicks } from './branches'
+import { CORNER, PAD, ROW_H, branchLabel, headLineage, layoutTree, momentTicks, savePointsUnder } from './branches'
 
 // The story tree, drawn.
 //
@@ -108,6 +108,9 @@ export function BranchMap({ branches, busyId, onSwitch, onRename, onDelete, onCl
   // would throw away the only copy of it.
   const save = async () => { if (await onRename(selected, renameText)) setRenameText(null) }
   const drop = async () => { if (await onDelete(selected)) setConfirming(false) }
+  // Save Points naming a moment on the selected branch, or on anything forked
+  // from it, keep it alive — the server refuses to delete them along with it.
+  const protecting = selected ? savePointsUnder(branches, selected.id) : 0
 
   return createPortal(
     <div className="modal-overlay branch-map-overlay" onClick={onClose}>
@@ -239,7 +242,13 @@ export function BranchMap({ branches, busyId, onSwitch, onRename, onDelete, onCl
 
             {confirming ? (
               <div className="branch-confirm">
-                <span>Delete this branch and everything forked from it?</span>
+                {/* The same wording the list gives, because the same deletion
+                    is reachable from both views and a rule that held in one of
+                    them would not be a rule. */}
+                <span>
+                  Delete this branch and everything forked from it? The story on
+                  other paths, and every Save Point, is unaffected.
+                </span>
                 <button type="button" className="danger" disabled={busy} onClick={drop}>
                   Delete
                 </button>
@@ -265,10 +274,14 @@ export function BranchMap({ branches, busyId, onSwitch, onRename, onDelete, onCl
                     the head stands on is offered, and says why it cannot go. */}
                 {!isRoot && (
                   <button type="button" className="danger"
-                    disabled={busy || isLoadBearing}
+                    disabled={busy || isLoadBearing || protecting > 0}
-                    title={isLoadBearing
+                    title={
-                      ? 'The line you are reading is built on this one. Switch away first.'
+                      isLoadBearing
-                      : undefined}
+                        ? 'The line you are reading is built on this one. Switch away first.'
+                        : protecting > 0
+                          ? `${protecting === 1 ? 'A Save Point is' : `${protecting} Save Points are`} saved on this line or one forked from it. Delete ${protecting === 1 ? 'it' : 'them'} first — deleting a Save Point deletes no story.`
+                          : undefined
+                    }
                     onClick={() => setConfirming(true)}>Delete</button>
                 )}
               </div>

frontend/src/branches.js

@@ -61,6 +61,41 @@ export function headLineage(branches) {
   return out
 }
 
+// How many Save Points would go if this branch were deleted.
+//
+// Deleting a branch deletes everything forked from it, and a Save Point names a
+// position on a line, so the Save Points on the whole doomed subtree go too.
+// The count is summed over that subtree rather than over the one branch — a
+// warning that said "1 Save Point" while three disappeared would be worse than
+// no warning at all.
+//
+// Computed here because the panel already holds every branch and its parent,
+// so the answer costs a walk rather than an endpoint. The server remains the
+// authority on what is actually deleted; this only lets the button say so
+// before it is pressed, the same division `headLineage` already uses.
+export function savePointsUnder(branches, rootId) {
+  const children = new Map()
+  for (const b of branches) {
+    const key = b.parent_branch_id
+    if (!children.has(key)) children.set(key, [])
+    children.get(key).push(b)
+  }
+  const byId = new Map(branches.map((b) => [b.id, b]))
+  const seen = new Set()
+  const stack = [rootId]
+  let total = 0
+  while (stack.length) {
+    const id = stack.pop()
+    // The guard is against a cycle, which the schema forbids and a walk should
+    // still never hang on.
+    if (seen.has(id)) continue
+    seen.add(id)
+    total += byId.get(id)?.save_points ?? 0
+    for (const child of children.get(id) ?? []) stack.push(child.id)
+  }
+  return total
+}
+
 // ---------- Map geometry ----------
 
 export const ROW_H = 64          // one branch, name above the lane and meta below

frontend/src/pages/Play/panels/BranchPanel.jsx

@@ -11,7 +11,7 @@
 import { useEffect, useState } from 'react'
 import { api } from '../../../api'
 import { BranchMap } from '../../../BranchMap'
-import { branchLabel, headLineage, orderBranches } from '../../../branches'
+import { branchLabel, headLineage, orderBranches, savePointsUnder } from '../../../branches'
 
 function BranchPanel({ advId, refreshKey, onSwitched, onTreeChanged, onError }) {
   const [branches, setBranches] = useState(null)
@@ -88,6 +88,12 @@ function BranchPanel({ advId, refreshKey, onSwitched, onTreeChanged, onError })
           // was forked from. The button said nothing about that and answered
           // with a toast; it now says so before it is pressed.
           const loadBearing = lineage.has(branch.id)
+          // A Save Point naming a moment on this branch, or on anything forked
+          // from it, keeps the branch alive: named moments last until the user
+          // deletes them, so the server refuses the deletion rather than taking
+          // them with it. Counted over the subtree, because deleting a branch
+          // takes its descendants.
+          const protecting = savePointsUnder(branches, branch.id)
           return (
             <div key={branch.id} className={`branch-row ${branch.is_head ? 'here' : ''}`}
               style={{ marginLeft: indent * 12 }}>
@@ -117,11 +123,20 @@ function BranchPanel({ advId, refreshKey, onSwitched, onTreeChanged, onError })
                 {branch.own_actions} of its own
                 {branch.parent_branch_id !== null && ` · forked at moment ${branch.fork_depth + 1}`}
                 {` · ends at ${branch.depth + 1}`}
+                {protecting > 0 &&
+                  ` · ${protecting === 1 ? '1 Save Point' : `${protecting} Save Points`} kept here`}
               </div>
 
               {isConfirming ? (
                 <div className="branch-confirm">
-                  <span>Delete this branch and everything forked from it?</span>
+                  {/* Only reachable when no Save Point points into this
+                      subtree — the button is disabled otherwise, and the server
+                      refuses regardless. So this says what the deletion costs
+                      without hedging about Save Points that cannot be at risk. */}
+                  <span>
+                    Delete this branch and everything forked from it? The story
+                    on other paths, and every Save Point, is unaffected.
+                  </span>
                   <button type="button" className="danger" disabled={busy}
                     onClick={() => remove(branch)}>Delete</button>
                   <button type="button" onClick={() => setConfirming(null)}>Keep</button>
@@ -145,10 +160,15 @@ function BranchPanel({ advId, refreshKey, onSwitched, onTreeChanged, onError })
                   {/* The root holds the turns every other branch borrows, and
                       the server refuses it — so it is not offered. */}
                   {branch.parent_branch_id !== null && (
-                    <button type="button" className="danger" disabled={busy || loadBearing}
+                    <button type="button" className="danger"
-                      title={loadBearing
+                      disabled={busy || loadBearing || protecting > 0}
-                        ? 'The line you are reading is built on this one. Switch away first.'
+                      title={
-                        : undefined}
+                        loadBearing
+                          ? 'The line you are reading is built on this one. Switch away first.'
+                          : protecting > 0
+                            ? `${protecting === 1 ? 'A Save Point is' : `${protecting} Save Points are`} saved on this line or one forked from it. Delete ${protecting === 1 ? 'it' : 'them'} first — deleting a Save Point deletes no story.`
+                            : undefined
+                      }
                       onClick={() => setConfirming(branch.id)}>Delete</button>
                   )}
                 </div>

planning/BROWSER-UX-SPEC.md

@@ -358,10 +358,16 @@ Each entry:
 
 ```text
 Before entering the abbey
-Turn 42
+Moment 42
 [Restore] [Rename] [Delete]
 ```
 
+**Moment, not Turn.** M4 closeout ruled in favour of the implemented product:
+the branch panel and the tree overlay already count in moments (`forked at
+moment 9`, `ends at moment 40`), so a Save Point list saying "Turn 42" would
+make one screen use two words for one thing. This is a vocabulary alignment and
+changes no behaviour; the number is unchanged.
+
 ## 26. Restore Confirmation
 
 Restoring is non-destructive.

planning/BUILD-MILESTONES.md

@@ -1,6 +1,6 @@
 # Adventure Storyteller — Production Build Milestones
 
-**Status:** In implementation. M1, M2 and M3 complete and accepted (M1 and M2: 2026-09-02; M3: 2026-09-03); M4 — Named Save Points / Checkpoints — implemented 2026-09-03, awaiting review  
+**Status:** In implementation. M1-M4 complete and accepted (M1 and M2: 2026-09-02; M3 and M4: 2026-09-03); M5 — Genre-Neutral Authoritative Narrative State — next to brief  
 **Base:** AI-DnD `d72f7c1bda0f34fccd84afb7a25c34eb01c901de`
 
 ## 1. Purpose
@@ -274,14 +274,18 @@ review. The architecture is recorded in **ADR 012**.
   a turn says while story descends from it off screen — both ratified in
   `STORY-BRANCH-SEMANTICS.md` (§5, §10, §14A).
 
-**Outstanding closeout condition:** the required **browser smoke test has not
+**Closeout condition — CLOSED at M4 closeout (2026-09-03).** M3 was accepted with
-been performed** — no session in which M3 was implemented or reviewed had a
+one condition outstanding: the required **browser smoke test had not been
-browser available. The equivalent sequence was driven end-to-end against the
+performed**, because no session in which M3 was implemented or reviewed had a
-running application with real inference and a process restart, and every
+browser available, leaving the DOM-level behaviour of the Redo button and its
-server-side behaviour it covers passes; the DOM-level behaviour of the Redo
+disabled states unverified by observation.
-button, its disabled states and its keyboard shortcut remain unverified by
+
-observation. This does not block M4, which touches none of that wiring, but it
+That condition is now discharged. A real Firefox 154.0.1, driven through
-remains an open M3 item until a human runs it.
+geckodriver, exercised M3's controls in the rendered application: Undo enabled
+and Redo disabled at the tip, two Undos moving the transcript back, Redo becoming
+enabled and returning the original tip exactly, Retry and the take pager, and a
+divergent write retiring Redo with no stale old-future text on screen. It passed.
+Evidence: `planning/reports/M4-IMPLEMENTATION-REPORT.md` §W.7.
 
 **Debt carried forward, none of it blocking M4:** full narrator-edit state
 re-evaluation is deferred to M5 (`STORY-BRANCH-SEMANTICS.md` §14A records the
@@ -351,12 +355,11 @@ M3's cost was reconciling them; a parallel checkpoint mover would recreate that
 divergence in a place where the two paths would silently disagree about what
 "restore" means. See ADR 012.
 
-## Status: IMPLEMENTED — awaiting review
+## Status: COMPLETE
 
-Implementation landed 2026-09-03. **Not accepted**: the milestone report has not
+Accepted 2026-09-03. Evidence: `planning/reports/M4-IMPLEMENTATION-REPORT.md`,
-been written and no reviewer has read the change. The Definition of Done above is
+including its §W closeout addendum. The Definition of Done is met, and — for the
-met by the code and the tests below; whether it is met by the *product* is what
+first time in this project — **verified in a real browser**.
-the review is for.
 
 **What M4 delivered:**
 
@@ -395,17 +398,77 @@ it.
 E-series lineage and memory isolation after restore and divergence, the edge
 cases in the brief, and the M3-database migration.
 
-**Outstanding condition, carried from M3 and not resolved here:** the **browser
+**Facts M5 inherits, and must not redesign:**
-smoke test has still not been performed**, for M3 or for M4. No session has had a
-usable browser. The M4 sequence was driven end-to-end over HTTP against a live
-server with a real process restart, and every server-side behaviour it covers
-passes; the DOM-level behaviour of the Save Point panel, its buttons and its
-confirmations remains unverified by observation.
 
-**Debt M4 carries forward:** none newly discovered in the head model. The Save
+- a Save Point is a **durable story coordinate** — `(branch, depth)` — carrying
-Point panel has no frontend test, because the project still has no frontend test
+  no copy of transcript, state, prompt, memory or summary;
-runner at all (M8). `POST /adventures/import` still returns every branch's rows
+- **restore is M3 head movement**, through the same `head.move_to` Undo and Redo
-rather than a head-capped window (inherited, M3).
+  use; there is no second restore path and M5 must not add one;
+- **state recovery stays snapshot/cached-position based**, never a replay of the
+  campaign (`TECHNICAL-DESIGN.md` §10.4). This is now load-bearing for Save
+  Points as well as Undo/Redo;
+- **the first divergent write** after a restore creates the continuation;
+  restore itself never forks;
+- **history a Save Point names cannot disappear** through an unrelated deletion
+  (§19.1);
+- **M3/M4 history and Save Point semantics are infrastructure now.** M5 replaces
+  the state *model*; it does not revisit how the story is positioned.
+
+**Acceptance evidence:** D11-D14, I04 and L03 all pass. D11 and L03 are
+discharged by automation that crosses a **genuine OS process boundary** — one
+server process writes the campaign, is killed, and a second process reads it back
+— rather than by recreating a client in one process.
+
+**The browser condition is closed, for M4 and retrospectively for M3.** A real
+Firefox 154.0.1, driven through geckodriver over the W3C WebDriver protocol,
+exercised the rendered DOM end to end: **44/44 checks passed**, covering M3's
+Undo/Redo enable states and transcript movement, Retry and the take pager, M3
+divergence and the disappearance of Redo, and every M4 Save Point operation
+including both confirmations and the branch-delete warning. No console errors.
+This closes the outstanding M3 condition recorded above and the equivalent M4
+one.
+
+**The three review findings were fixed during closeout:**
+
+- **B-1** — the Save Point list was an N+1 that loaded whole `Action` rows,
+  narration included. It is now one bulk two-column coordinate query plus one
+  lineage: **53 SELECTs for 25 Save Points became 5**, and the query count no
+  longer moves with the length of the list.
+- **B-2** — deleting a branch silently deleted the Save Points naming it. Fixed
+  as a **behaviour** defect, not a wording one: a branch a Save Point names can
+  no longer be deleted at all. The request is refused with the offending Save
+  Points named, the user deletes them explicitly (which deletes no story), and
+  the branch then goes. Both delete controls, in the branch list and in the tree
+  overlay, disable and explain rather than warning about a loss that no longer
+  happens. `STORY-BRANCH-SEMANTICS.md` **§19.1** records the rule; §28 already
+  required a future cleanup feature to retain checkpoint-referenced paths, and
+  this is that requirement applied to the deletion path that exists today.
+- **B-3** — the D11/L03 automation now spawns real server processes.
+
+**Also fixed:** creating a Save Point takes the campaign's turn lock (review §S
+C-5), so "save where I am" cannot read a head that a turn in flight is about to
+move. Rename and Delete deliberately do not take it — neither reads nor moves a
+story position.
+
+**Debt M4 carries forward:** the Save Point panel has no frontend test, because
+the project still has no frontend test runner at all (M8) — the browser smoke
+test above is a closeout procedure, not a suite. `POST /adventures/import` still
+returns every branch's rows rather than a head-capped window (inherited, M3).
+
+## Note to M5 — the instrumentation is now larger than M3 estimated
+
+M4 added **60 tests that use inherited RPG world-state values as deterministic
+instrumentation**, on top of the ~20 M3 flagged. M5 replaces that state model,
+and must **move the instrumentation while preserving the behavioural
+assertions**: what those tests measure is where the story is being read and what
+state belongs to that position, which is exactly as true after M5 as before it.
+Deleting them would delete the evidence for D11-D14, I04, L03 and the E-series.
+
+The existing constraint stands and is now load-bearing for Save Points as well as
+Undo/Redo: **historical state must remain efficiently snapshot/cache
+recoverable**, so that moving to a position never becomes proportional to
+campaign length (`TECHNICAL-DESIGN.md` §10.4). Restore, Undo and Redo all pay
+whatever that costs.
 
 ---
 

planning/DATA-MODEL.md

@@ -230,6 +230,30 @@ It carries **no** copy of the transcript, the state, the prompt, a memory, a
 summary, or a branch's contents. Everything a restore produces comes from the
 retained history the coordinate points into.
 
+The retry case is what settles the coordinate-versus-turn-id question, and M4
+closeout measured it rather than arguing it. A Save Point named a turn whose
+live row was id 17; retrying that turn made id 17 dead and id 18 live at the
+same coordinate; the Save Point resolved to id 18 and restored correctly. A row
+id would have pinned a take the story no longer tells. **A Save Point names a
+story position, not a particular take of it.**
+
+Three further properties of the implemented model, recorded so they are decided
+rather than incidental:
+
+- **Names are not unique**, and nothing requires them to be. No product
+  requirement asks for uniqueness, and two names for one moment is a reasonable
+  thing for a player to want.
+- **Several Save Points may name the same position.** Same reason.
+- **Ordinary list presentation is newest-created first.** Story order is not
+  something the list can honestly claim: depths on lines that have parted
+  company are not comparable, so ordering by depth would draw a sequence that no
+  reading of the story passes through. When each was made is a fact about all of
+  them.
+
+None of this is genre-specific. A coordinate is a position in a story; what the
+state at that position *contains* is M5's question, and changing it does not
+change what a Save Point is.
+
 Two consequences worth recording here:
 
 - **Restore reuses the campaign's one head-movement mechanism.** It resolves the
@@ -238,11 +262,13 @@ Two consequences worth recording here:
   the path being read, which is what makes a Save Point on a departed line
   restorable at all — and what keeps a Save Point in a shared prefix from
   dragging the reader off the line they chose. See `TECHNICAL-DESIGN.md` §8.8.
-- **A checkpoint is durable against everything but its own deletion and its
+- **A checkpoint is durable against everything but its own explicit deletion.**
-  branch's.** No pass removes one for going stale, sitting behind the head, or
+  No pass removes one for going stale, sitting behind the head, or naming a line
-  naming a line the story left (`STORY-BRANCH-SEMANTICS.md` §19). Deleting a
+  the story left (`STORY-BRANCH-SEMANTICS.md` §19). Deleting a *branch* does not
-  branch removes its checkpoints by cascade, as it removes its memories, because
+  remove one either: the deletion is refused while a checkpoint names any
-  the story they named is gone.
+  position in the subtree, and the user deletes the checkpoint first
+  (§19.1). Deleting the whole campaign removes them, which is what deleting a
+  campaign means.
 
 ## 9. Narrative Entity
 

planning/PROJECT-SOURCES.md

@@ -85,13 +85,13 @@ One file, and it changes as development progresses:
 planning/reports/M4-IMPLEMENTATION-REPORT.md
 ```
 
-M3 is the most recently completed milestone, and M4 is the next to be briefed.
+M4 is the most recently completed milestone, and M5 is the next to be briefed.
-This report is M3's review *and* its primary evidence record — no separate M3
+This report is M4's review *and* its closeout record: its §W holds the corrective
-baseline report was produced — so it is the only place some of what M3 left
+work, the process-boundary automation, and the real-browser verification that
-behind is written down, including the browser smoke test M3 still owes.
+closed the condition M3 and M4 both carried.
 
-**Replace it, do not accumulate.** When M4's report lands, remove this one from
+**Replace it, do not accumulate.** When M5's report lands, remove this one from
-the project Sources and upload M4's instead. The repository does the same thing:
+the project Sources and upload M5's instead. The repository does the same thing:
 `planning/reports/` holds the current milestone's report and
 `planning/archive/milestone-reports/` holds the rest.
 

planning/README.md

@@ -3,10 +3,10 @@
 **This file is the index. Start here.**
 
 **Current state:** Phase 0 complete; AI-DnD forked as the production base;
-milestones **M1, M2 and M3 implemented and accepted** (M3: 2026-09-03).
+milestones **M1, M2, M3 and M4 implemented and accepted** (M3 and M4:
-**M4 — named Save Points — is implemented (2026-09-03) and awaiting review.**
+2026-09-03).
-Its implementation report has not been written, and writing it is the current
+**Next:** **M5 — Genre-Neutral Authoritative Narrative State.** Its brief has not
-action. Do not begin M5.
+been written yet, and writing it is the current action.
 
 **Package version:** see `VERSION.md`, which records what each revision changed
 and why.
@@ -234,9 +234,9 @@ Milestone M3                        COMPLETE (2026-09-03)
   active-head export                and ADR 012
         |
         v
-Milestone M4                        IMPLEMENTED 2026-09-03 —
+Milestone M4                        COMPLETE (2026-09-03)
-  named Save Points                 awaiting review; no report yet
+  named Save Points                 reports/M4-IMPLEMENTATION-REPORT.md
-        |
+        |                           browser verification: PASS (M3 + M4)
         v
 M5-M11, one at a time               see BUILD-MILESTONES.md
 ```
@@ -245,16 +245,17 @@ M5-M11, one at a time               see BUILD-MILESTONES.md
 
 **One milestone at a time. Do not begin a milestone before its brief exists.**
 
-**M4 is implemented and unreviewed.** Its implementation report is the current
+**No M5 brief has been prepared.** Writing one is the current action, informed by
-action; M5 does not begin before that report is written and accepted.
+the M4 report's §U readiness assessment and by the note `BUILD-MILESTONES.md`
+attaches to M5 — in particular that M4 added 55 tests using the inherited
+world-state values as deterministic instrumentation, which M5 must **move rather
+than delete**.
 
-Two conditions remain open. The **browser smoke test has still not been
+**No conditions remain open on M1-M4.** The browser smoke condition that M3 and
-performed** — now for M3 and for M4 — because no session so far has had a usable
+M4 both carried was satisfied at M4 closeout: a real Firefox exercised the
-browser. See `archive/milestone-reports/M3-IMPLEMENTATION-REPORT.md` §M and §W.4,
+rendered DOM for both milestones' controls, 44/44 checks passing. The M3 report's
-`reports/M4-IMPLEMENTATION-REPORT.md` §M, and the M4 status block in
+§M.2 and the M4 report's §M record the condition as it stood; the M4 report's §W
-`BUILD-MILESTONES.md`. And **no M4 review exists**: the status block was
+records it closed.
-written by the implementation and records what it built, which is not the same as
-a reviewer having read it.
 
 ## What each milestone closeout corrected
 

planning/STORY-BRANCH-SEMANTICS.md

@@ -467,6 +467,39 @@ They should not be removed by:
 - branch divergence,
 - ordinary history cleanup.
 
+### 19.1 A checkpoint protects the history it names
+
+Added at M4 closeout, from implementation evidence.
+
+§19's durability rule is only meaningful if something enforces it against the
+operations that delete history. Deleting a branch deletes that branch and
+everything forked from it, so a checkpoint naming a position anywhere in that
+subtree would go with it — and go silently, because the story is what the user
+asked to delete and the named moments are not mentioned in the request.
+
+Therefore:
+
+- **A branch cannot be deleted while a checkpoint names a position on it, or on
+  any branch forked from it.** The deletion is refused, not amended: nothing is
+  half-deleted, and no checkpoint is quietly relocated or dropped.
+- **The refusal names the checkpoints standing in the way**, so the user knows
+  what to act on rather than being told only that something is in the way.
+- **The user resolves it by deleting the checkpoint explicitly**, which is
+  §19's "until explicitly deleted" being honoured rather than worked around.
+  Deleting a checkpoint still deletes no story (§25), so the recovery costs the
+  user nothing they wanted to keep.
+- **The scope is the subtree**, not the named branch alone.
+
+This is the same rule §28 already anticipates for a future cleanup feature —
+*retain paths referenced by checkpoints* — applied to the one deletion path that
+exists today. A cleanup or discarded-history feature added later must honour it
+too, and must not acquire a way to delete a checkpoint as a side effect of
+deleting history.
+
+Campaign deletion is not an exception to this and needs no rule: deleting a
+campaign deletes everything in it, checkpoints included, and that is what the
+user asked for.
+
 ## 20. Checkpoint Restore
 
 Restoring a checkpoint:

planning/TECHNICAL-DESIGN.md

@@ -452,9 +452,17 @@ once they have.
 
 **Nothing removes a Save Point but the user.** There is no cleanup pass, and none
 is wanted: a Save Point pointing behind the head, or into a line the story left,
-is doing its job. The one exception is referential and not a policy — deleting a
+is doing its job.
-branch takes its Save Points with it, by the same cascade that takes its
+
-memories, because the story they named went with it.
+That rule is enforced against the one operation that could break it. Deleting a
+branch deletes everything forked from it, so a Save Point naming a position in
+that subtree would go too — silently, since the story is what the user asked to
+delete. **The deletion is therefore refused while any Save Point names that
+subtree**, and the refusal names them. The user deletes the Save Point
+explicitly, which deletes no story, and then the branch. `STORY-BRANCH-SEMANTICS.md`
+§19.1 states the rule; §28 already required a future cleanup feature to retain
+paths a checkpoint references, and this is that requirement applied to the
+deletion path that exists today.
 
 ## 9. Export / Import and Head Position
 

planning/V1-ACCEPTANCE-TESTS.md

@@ -1,8 +1,28 @@
 # Adventure Storyteller — V1 Acceptance Tests
 
-**Status:** v1.2 planning/release contract — updated after Phase 0B, after M2 for
+**Status:** v1.3 planning/release contract — updated after Phase 0B, after M2 for
-the security contract (H10 strengthened, H12 added), and after M3 for history
+the security contract (H10 strengthened, H12 added), after M3 for history
-ownership and results (D03, D10, I07, L01)  
+ownership and results (D03, D10, I07, L01), and after M4 for Save Point results
+(D11-D14, I04, L03, E-series) and the browser condition below  
+
+> **Browser-level verification (M4 closeout, 2026-09-03).** The browser smoke
+> condition that M3 and M4 both carried is **satisfied**. A real Firefox 154.0.1,
+> driven through geckodriver over the W3C WebDriver protocol, exercised the
+> rendered DOM: M3's Undo/Redo enable states, transcript movement, Retry and the
+> take pager, divergence and the loss of Redo; and M4's full Save Point
+> lifecycle including both confirmations and the branch-delete warning. 44/44
+> checks passed with no console errors, on two independent runs. No pass
+> condition anywhere in this document was changed to achieve it. See
+> `reports/M4-IMPLEMENTATION-REPORT.md` §W.
+>
+> **Three kinds of evidence are recorded separately below, and are not
+> interchangeable.** *Automated* means a test in the repository's suite, which
+> runs on every future change. *Live runtime* means a real server exercised over
+> HTTP — stronger than a unit test about process boundaries, weaker than a
+> browser about anything a user sees. *Browser* means the rendered DOM driven by
+> a real browser, which is the only evidence that a control is visible, enabled
+> and wired. Where a result cites more than one, the strongest is named last.
+
 **Purpose:** Define black-box acceptance tests for finalist evaluation during Phase 0B and for the eventual v1 release.
 
 ## 1. Test Philosophy
@@ -737,6 +757,14 @@ Before entering the abbey
 ### Pass
 Checkpoint persists across application restart.
 
+### Result — PASS (M4 closeout, 2026-09-04)
+*Automated:* `backend/tests/test_process_restart.py` starts the application as a
+subprocess, writes the campaign, **terminates the process**, and starts a second
+process against the same database — the Save Point, its name and its
+`(branch, depth)` coordinate all survive.
+*Browser:* the Save Point is still listed after a full page reload
+(`reports/M4-IMPLEMENTATION-REPORT.md` §W.7 section E).
+
 ---
 
 ## D12 — Restore Checkpoint
@@ -751,6 +779,11 @@ Checkpoint persists across application restart.
 ### Pass
 Transcript/state return to checkpoint position.
 
+### Result — PASS (M4 closeout, 2026-09-04)
+*Automated:* `test_d12_restore_returns_the_transcript_and_the_state`.
+*Browser:* the visible transcript and the state both move back, and the view
+refreshes without a manual reload (§W.7 section F).
+
 ---
 
 ## D13 — Restore Does Not Delete Later History
@@ -760,6 +793,13 @@ Transcript/state return to checkpoint position.
 ### Pass
 Later story is retained as abandoned/disposable history.
 
+### Result — PASS (M4 closeout, 2026-09-04)
+*Automated:* measured on **row identity**, not on counts — the set of action row
+ids before a restore equals the set after it. Ordinary Redo still walks the retained
+continuation until a divergent write, and after that write the displaced rows are
+still present while Redo reports nothing ahead. Four `test_d13_*` tests, and
+confirmed in the browser with a database check behind it.
+
 ---
 
 ## D14 — Delete Checkpoint
@@ -773,10 +813,31 @@ Delete named checkpoint.
 - checkpoint pointer disappears,
 - referenced story turn/history remains intact.
 
+### Result — PASS (M4 closeout, 2026-09-04)
+*Automated:* the pointer row goes; the referenced turn, the later history and the
+active head are all unchanged (`test_d14_delete_removes_the_pointer_and_no_story`).
+*Browser:* the confirmation states that deleting the Save Point does not delete
+the story, and the story remains afterwards (§W.7 section I).
+
+A Save Point is also the **only** thing that can remove itself: deleting a branch
+whose history a Save Point names is refused rather than cascading
+(`STORY-BRANCH-SEMANTICS.md` §19.1), verified automatically and in the browser.
+
 ---
 
 # E. Branch and Lineage Safety
 
+> **M4 result (2026-09-03).** E01 and E04 were re-exercised through a Save Point
+> restore rather than only through Undo, and pass: a memory derived past a
+> restored head stops being retrievable and becomes eligible again on Redo,
+> without being deleted or re-embedded; after restore-plus-divergence the old
+> future's memory stays ineligible even as the new line grows past its depth; and
+> the transcript after a restore holds only the active lineage while the
+> displaced rows remain in the tree. **E03** (summary lineage over a long story)
+> remains **NOT PERFORMED** — it needs a long-run campaign and is owned by
+> M6/M11, unchanged from M3.
+
+
 ## E01 — Abandoned Future Cannot Affect Active State
 
 **Priority:** REQUIRED FOR V1
@@ -1353,6 +1414,15 @@ Export preserves retained alternate/disposable history needed for recovery, unle
 ### Pass
 Named checkpoints survive export/import.
 
+### Result — PASS (M4 closeout, 2026-09-04)
+*Automated and live runtime.* A real round trip with three Save Points across two
+branches: names, notes and
+coordinates survive, branch references are remapped to the imported rows
+(1→3, 2→4), and each restores to a distinct position and state in the new
+campaign. Importing Save Points does **not** move the active head — the head
+still comes from the bundle's `headDepth`. Bundles written before M4 carry no
+`checkpoints` key, import cleanly, and create none.
+
 ---
 
 ## I05 — Knowledge Provenance Export
@@ -1581,6 +1651,14 @@ State at each position matches original accepted state.
 ### Pass
 Correct historical state is reconstructed.
 
+### Result — PASS (M4 closeout, 2026-09-04)
+*Automated, across a genuine OS process boundary:* the state at the Save Point
+was recorded before the first process exited, and a second process restored
+exactly that value after the campaign had been advanced past it
+(`backend/tests/test_process_restart.py`). This replaced a same-process
+`TestClient` restart, which could not distinguish durable state from a live
+object.
+
 ---
 
 ## L04 — Derived Data Can Be Rebuilt

planning/VERSION.md

@@ -1,8 +1,93 @@
 # Planning Package Version
 
-- **Package:** Adventure Storyteller Planning Package v2.5
+- **Package:** Adventure Storyteller Planning Package v2.6
 - **Revision date:** 2026-09-03
-- **Status:** Phase 0 complete; architecture selected; **Milestones M1, M2 and M3 implemented and accepted**; **M4 implemented 2026-09-03 and awaiting review.**
+- **Status:** Phase 0 complete; architecture selected; **Milestones M1-M4 implemented and accepted**; M5 is next to brief.
+
+## v2.6 — M4 Closeout (2026-09-03)
+
+M4 is **accepted**. Its review returned *PASS WITH CORRECTIVE WORK REQUIRED*; the
+corrective work is done, and the browser condition that M3 and M4 both carried is
+closed.
+
+**The three review findings, fixed:**
+
+- **B-1 — the Save Point list was an N+1.** It resolved each Save Point with its
+  own query and loaded whole `Action` rows, narration included, to answer "does a
+  row exist here". It is now one bulk two-column coordinate query plus one
+  lineage computation: **53 SELECTs for 25 Save Points became 5**, and the count
+  no longer grows with the list. Guarded by three tests, including one proving
+  the coordinate is matched as a *pair* — an `IN`-list version would report a
+  Save Point resolved because another branch has a live row at the same depth.
+- **B-2 — deleting a branch silently deleted its Save Points.** Fixed as a
+  **behaviour** defect rather than a missing warning, because
+  `STORY-BRANCH-SEMANTICS.md` §19 says a named checkpoint remains until
+  explicitly deleted and §28 already required future cleanup to retain
+  checkpoint-referenced paths. **A branch a Save Point names can no longer be
+  deleted.** The request is refused with the offending Save Points named; the
+  user deletes them explicitly, which deletes no story, and the branch then
+  goes. Both delete controls disable and explain. Recorded as a new
+  **§19.1**.
+- **B-3 — the D11/L03 automation never left one process.** A new module spawns
+  real server processes, kills the first, and reads the campaign back with the
+  second.
+
+**Also fixed (review §S C-5):** creating a Save Point now takes the campaign's
+turn lock, so "save where I am" cannot read a head a turn in flight is about to
+move. Rename and Delete deliberately do not take it, and a test pins that
+decision.
+
+**Real-browser verification — the first in this project.** A Firefox 154.0.1
+driven through geckodriver over the W3C WebDriver protocol exercised the rendered
+DOM for **both** milestones: **44/44 checks passed**, no console errors. It
+covered M3's Undo/Redo enable states, transcript movement, Retry and the take
+pager, and divergence retiring Redo; and M4's whole Save Point lifecycle
+including both confirmations and the new branch-delete warning. **The outstanding
+M3 browser condition is therefore closed as well.** No dependency was added to
+the repository: the WebDriver client for the run was written against stdlib HTTP.
+
+**Also corrected, found while fixing B-2:** `models.py`, `TECHNICAL-DESIGN.md`
+§8.8 and `DATA-MODEL.md` §8 all described the cascade as the durability rule.
+They now describe the refusal, and record that `checkpoints.branch_id`'s cascade
+survives as referential integrity that the application no longer reaches.
+
+**Documents corrected by this closeout:**
+
+- `V1-ACCEPTANCE-TESTS.md` records results for **D11-D14, I04, L03** and the
+  E-series, and states that the browser-level condition is satisfied. **No pass
+  condition was weakened** — and D11/L03 now note that the automation crosses a
+  genuine OS process boundary, which is the standard later milestones should
+  meet.
+- `DATA-MODEL.md` §8 records the coordinate as implemented, with the retry
+  measurement that settles coordinate-versus-turn-id, and three decisions that
+  were previously implicit: names are not unique, several Save Points may name
+  one position, and the list is newest-created first.
+- `STORY-BRANCH-SEMANTICS.md` gains **§19.1** — a checkpoint protects the
+  history it names. This is the only behavioural specification change in the
+  closeout, and it strengthens §19 rather than weakening anything.
+- `BROWSER-UX-SPEC.md` §25 rules for the implemented vocabulary: **Moment N**,
+  not *Turn N*, because the branch panel and tree overlay already count in
+  moments. Vocabulary only; no behaviour changes.
+- `BUILD-MILESTONES.md` marks **M4 COMPLETE**, records the fixes and the browser
+  result, and warns M5 that the instrumentation to move is now **55 tests**.
+- `README.md` records M1-M4 accepted and M5 as next to brief.
+- `PROJECT-SOURCES.md` and `project-sources.txt` point at the current report;
+  `project-sources.txt` still named the archived M3 report and was corrected.
+
+**Report rotation** happened in the reporting pass that preceded this closeout:
+`M3-IMPLEMENTATION-REPORT.md` moved to `archive/milestone-reports/` as a pure
+rename, and `reports/` now holds M4's report, whose **§W** is this closeout's
+evidence record.
+
+**No new ADR.** ADR 012 already decides the architecture, and the corrective work
+forced no new architectural decision. `SPECIFICATION.md`,
+`STORY-BRANCH-SEMANTICS.md`, `SECURITY-THREAT-MODEL.md`, `CONTEXT-AND-MEMORY.md`,
+`IMPORTED-KNOWLEDGE-DESIGN.md` and ADRs 003, 005 and 012 are unchanged.
+
+**M5 readiness:** ready. Save Points store no state and no checkpoint code reads
+any, so M5 can change what a snapshot contains without touching what a Save Point
+is — provided it keeps state recoverable at a position without replay
+(`TECHNICAL-DESIGN.md` §10.4).
 
 ## v2.5 — M4 Implementation (2026-09-03)
 

planning/project-sources.txt

@@ -26,4 +26,4 @@ planning/DECISIONS/009-ai-dnd-production-base.md
 planning/DECISIONS/010-explicit-typed-narrative-state-events.md
 planning/DECISIONS/011-local-inference-endpoint-policy.md
 planning/DECISIONS/012-active-head-non-destructive-history.md
-planning/reports/M3-IMPLEMENTATION-REPORT.md
+planning/reports/M4-IMPLEMENTATION-REPORT.md

planning/reports/M4-IMPLEMENTATION-REPORT.md

@@ -1040,4 +1040,279 @@ second consecutive milestone. What to do with that is the reviewer's call.
 
 ---
 
+# W. M4 Closeout Addendum (2026-09-03 / 2026-09-04)
+
+**This addendum is current where it differs from the review above**, and §W.3 is
+current where it differs from an earlier draft of this addendum: B-2 was first
+addressed as a warning over a surviving cascade, then reclassified as a behaviour
+defect and fixed by refusing the deletion. Only the final behaviour is described
+below; the review's §R still records the finding as it was raised. Sections
+A-V record what was true at commit `e08d49c`, before corrective work; nothing in
+them has been rewritten. Where they say a finding is open or a test is
+unperformed, this section says what happened next.
+
+## W.1 Starting state
+
+| Fact | Value |
+| --- | --- |
+| Closeout started from | `279a871` — *Planning: add the M4 implementation review report and rotate M3's* |
+| Closeout dates | corrective work and browser verification 2026-09-03; B-2 reclassified and re-fixed 2026-09-04 |
+| Reviewed implementation | `e08d49c` — signed, unchanged by this pass |
+| Working tree at start | clean; the report and rotation were already committed |
+| Upstream ancestry | intact |
+| LICENSE | unchanged |
+
+## W.2 B-1 — the Save Point list N+1 is gone
+
+`GET /checkpoints` resolved each Save Point with its own query and loaded whole
+`Action` entities to do it. It now takes one bulk query over two integer columns
+plus one lineage computation, both before the loop.
+
+Measured on the same 25-Save-Point fixture the review used:
+
+| | before | after |
+| --- | --- | --- |
+| SQL statements per list | **53** | **5** |
+| SELECTs against `actions` | 25 | **1** |
+| columns in that SELECT | 11, including `text` | **2** (`branch_id`, `depth`) |
+| narration fetched | yes | **no** |
+| queries per Save Point | 2.1 | 0.2 |
+
+Three tests guard it. One asserts on *growth* rather than a fixed budget — five
+times the Save Points must not mean more queries — because a fixed number is
+something to edit rather than a rule to keep. One asserts the emitted SQL never
+names `actions.text`, `actions.reasoning` or `actions.world_delta`. The third is
+the one worth keeping: it builds a **cross-product trap**, a dead coordinate on
+one branch while another branch has a live row at the same depth, and proves the
+dead one still reports `resolved: false`. An `IN`-list implementation would fail
+it; the OR-of-pairs passes.
+
+`_render_all` is now the single rendering path — create and rename call it
+through `_rendered` with a list of one — so the list and the single-item
+responses cannot drift.
+
+## W.3 B-2 — a branch a Save Point names cannot be deleted
+
+The review recorded this as a missing warning. On closeout it was reclassified as
+a **behaviour defect**, because the authority is unambiguous:
+`STORY-BRANCH-SEMANTICS.md` §19 says a named checkpoint remains until explicitly
+deleted, and §28 already requires a future cleanup feature to *retain paths
+referenced by checkpoints*. A cascade that removed Save Points along with a
+branch violates both, and a warning would only have documented the violation.
+
+**The deletion is now refused.** `DELETE /branches/{id}` returns **409** when any
+Save Point names a position on that branch or on anything forked from it — the
+subtree, because deleting a branch takes its descendants, and a check scoped to
+the named branch alone would let a Save Point on a child vanish silently.
+
+The refusal names what stands in the way rather than only counting it:
+
+```text
+This branch, or a branch forked from it, is where a Save Point “On the new
+line” is saved. Delete that Save Point first if you no longer need it, then
+delete the branch. Deleting a Save Point does not delete any story.
+```
+
+Long lists truncate ("and 2 more") so the message stays a sentence someone reads.
+
+The recovery is the point, and it is cheap: deleting a Save Point deletes no
+story (§25), so the user removes the pointer and the branch then goes. Verified
+end to end in the browser (§W.7 section J).
+
+Both delete controls — the branch list and the tree overlay — now **disable**
+Delete while Save Points protect the subtree and say why, and the branch row
+reports `· 2 Save Points kept here`. The confirmation, now only reachable when
+nothing is at risk, says plainly that the story on other paths and every Save
+Point are unaffected. Two views changed, because the same deletion is reachable
+from both and a rule holding in one of them would not be a rule.
+
+`checkpoints.branch_id` keeps `ON DELETE CASCADE` as referential integrity — a
+Save Point must never point at a branch that is gone — but the guard means it
+does not fire through the application. Campaign deletion still cascades, which is
+what deleting a campaign means.
+
+Six tests cover the five required scenarios plus the truncation: deletion refused
+and nothing lost; the Save Point intact after the refusal; explicit deletion then
+allowing the branch delete; a branch no Save Point names still deleting; a Save
+Point on a *descendant* also protecting; and an unrelated Save Point not blocking
+anything. A source-level test asserts both views disable and explain.
+
+**Documents this corrected beyond the planned set:** `models.py`,
+`TECHNICAL-DESIGN.md` §8.8 and `DATA-MODEL.md` §8 had all recorded the cascade as
+the durability rule. They now record the refusal.
+
+## W.4 B-3 — the restart test now crosses a process boundary
+
+`backend/tests/test_process_restart.py` (3 tests, ~9 s) starts the real
+application with `subprocess.Popen`, plays a story over HTTP, **kills the
+process**, starts a second process against the same database file, and only then
+asks its questions. Readiness is probed, never slept on; children are terminated
+in a `finally` whether or not the test passes.
+
+It covers D11 (the Save Point, its name and its coordinate survive), L03 (the
+state at the position returns), that restore deletes no rows, that Redo is
+available before divergence, that the retained continuation is still walkable,
+and that a divergent write after a restart still forks on the write rather than
+the restore.
+
+The old same-process helper was **not** kept as a pretend equivalent.
+
+## W.5 C-5 — creating a Save Point is serialized
+
+Create now takes the campaign's existing turn lock, the same one Undo, Redo and
+Restore take. No new lock was introduced. "Save where I am" has to name one
+committed position, and the head is exactly what a turn in flight is about to
+move.
+
+Rename and Delete deliberately **do not** take it: neither reads nor moves a
+story position, and refusing a label edit during generation would be a worse
+product for no safety gained. A test pins that decision so it reads as a choice
+rather than an oversight, and another proves a refused create releases the lock.
+
+## W.6 Test and build results
+
+All against the closeout tree.
+
+| Suite | Result |
+| --- | --- |
+| **Full backend** | **698 passed**, 0 failed, 0 skipped (was 680) |
+| **M4 targeted** (`test_save_points.py` + `test_process_restart.py`) | **60 passed** (57 + 3) |
+| **M3 invariants** (8 modules) | **130 passed** |
+| **Security / local-only** (incl. `test_egress.py`) | **93 passed** |
+| **Frontend lint** | exit 0; 7 warnings, unchanged from baseline |
+| **Frontend build** | exit 0 |
+| **Docker build** | exit 0 |
+
+698 − 680 = 18 new tests: 3 process-restart, 4 for B-1, 7 for B-2 (the refusal
+rule and its recovery), 3 for C-5, and one holding the branch-list count to the
+same subtree the server refuses on. Two tests written earlier in this closeout
+were **replaced**, not kept alongside: they asserted the cascade behaviour that
+§W.3 removed, and leaving them would have pinned the defect.
+
+## W.7 Browser verification — PERFORMED, and it passes
+
+**This is the first real-browser verification in the project**, and it discharges
+the condition M3 and M4 both carried.
+
+| | |
+| --- | --- |
+| Browser | **Mozilla Firefox 154.0.1**, headless |
+| Driver | geckodriver 0.37.1, W3C WebDriver over HTTP |
+| Client | written for the run against Python's stdlib — **no dependency added to the repository** |
+| App | the real production-shaped server, SPA served same-origin, deterministic scripted model |
+| Result | **47/47 checks passed**, no console errors; run twice on independent fresh databases |
+
+§M reported no usable browser, and that was true of the paths tried there: the
+`firefox` snap wrapper fails with mount-namespace errors and hangs on a headless
+screenshot. The binary **inside** the snap
+(`/snap/firefox/current/usr/lib/firefox/firefox`) runs correctly under
+geckodriver, which §M did not try. §M's conclusion is superseded; its account of
+what was attempted stands.
+
+What the browser actually did, by section of the closeout brief:
+
+| Section | Verified in the DOM |
+| --- | --- |
+| **A — M3 history** | transcript renders; Undo enabled and Redo disabled at the tip; two Undos move the transcript back twice; Redo becomes enabled; two Redos return the original tip exactly |
+| **B — retry / takes** | Retry produces an alternate take; the take pager appears; stepping between takes changes the visible narration; no branch vocabulary needed |
+| **C — M3 divergence** | a new continuation from a moved-back head appears; Redo becomes unavailable; no stale old-future text in the active transcript; position and continuation survive a reload — with a database check confirming the displaced rows are still on disk |
+| **D — creation** | Save Point created and named through the form; appears in the list; panel says **Save Point** with no branch/head/node/fork wording; position reads **Moment N** |
+| **E — persistence** | still present after a full page reload |
+| **F — restore** | the confirmation visibly explains later history is kept; transcript and state move back; the Save Point stays listed; Redo becomes available; zero rows deleted (database check); the view refreshes without a manual reload |
+| **G — restore + redo** | Redo returns the original continuation; the Save Point survives |
+| **H — restore + divergence** | the different continuation appears; Redo into the displaced future is gone; no old-future narration in the active view; the displaced rows retained on disk |
+| **I — rename / delete** | rename changes the name and not the position; it still restores to the same moment; the delete warning says the story is not deleted; the row disappears with no stale list; the story remains |
+| **J — branch deletion** | the branch row reports the Save Points kept on it; Delete is **disabled** and explains what to do; the server independently refuses with **409** naming the Save Point; nothing is deleted by the refusal; and the documented recovery works — deleting the Save Point frees the branch |
+| **K — control state** | enable/disable states match the position; no material console errors |
+
+**No defect was found in the application by the browser run.** Four failures
+occurred and all four were in the harness: a wrong SPA route (`/adventures/:id`
+rather than `/play/:id`); a wait predicate that compared transcript *length* when
+the empty-story placeholder is longer than the first turn; a fixture that tried
+to delete the branch it was reading, which is refused by design; and a reload
+assertion that sampled the transcript once instead of waiting for it to render.
+The last was checked against the application before being called a harness bug —
+after a reload the text is present at the first sample, so the race was the
+test's.
+
+## W.8 Planning documents updated
+
+`V1-ACCEPTANCE-TESTS.md` (D11-D14, I04, L03, E-series results and the browser
+condition; **no pass condition weakened**), `STORY-BRANCH-SEMANTICS.md` **§19.1**
+(new — a checkpoint protects the history it names; the closeout's only
+behavioural specification change, and it strengthens §19),
+`DATA-MODEL.md` §8, `TECHNICAL-DESIGN.md` §8.8, `BROWSER-UX-SPEC.md` §25 (Moment,
+not Turn), `BUILD-MILESTONES.md` (M4 COMPLETE, the facts M5 inherits, the M5
+instrumentation note), `planning/README.md`, `VERSION.md` (v2.6),
+`PROJECT-SOURCES.md` and `project-sources.txt`, which still named the archived M3
+report.
+
+Unchanged, as §T recommended: `SPECIFICATION.md`, `STORY-BRANCH-SEMANTICS.md`,
+`SECURITY-THREAT-MODEL.md`, `CONTEXT-AND-MEMORY.md`,
+`IMPORTED-KNOWLEDGE-DESIGN.md`, ADR 003, ADR 005, ADR 012. **No ADR 013**: the
+corrective work forced no architectural decision. §19.1 is a durability rule
+inside an existing specification, not a new architecture — ADR 005 already says
+history is preserved rather than overwritten, and refusing to delete a
+checkpoint's history is that decision applied, not a departure from it.
+
+## W.9 The architecture is unchanged
+
+Worth stating plainly, because closeout touched the checkpoint module. A Save
+Point is still `name + optional note + (branch, depth)`. Restore is still
+`coordinate → head.move_to_node → head.move_to → attempts.restore_state`. The
+corrective work touched **how the list is read** and **when create is allowed to
+read the head** — never what a Save Point is or how restoring one moves the
+story. Nothing was copied into a checkpoint, no checkpoint-specific Redo stack
+exists, restore still does not fork, and the first divergent write is still what
+creates the continuation.
+
+## W.10 Remaining debt
+
+Unchanged from §S and none of it M4's: no frontend test runner (M8) — the browser
+run above is a closeout procedure, not a suite; `POST /adventures/import` returns
+every branch's rows rather than a head-capped window (inherited, M3); the RPG
+world-state instrumentation, now **60 tests**, moves at M5.
+
+## W.11 Result
+
+D11, D12, D13, D14, I04 and L03 all **PASS**. The E-series lineage results
+through a Save Point restore all **PASS** (E03 remains NOT PERFORMED, owned by
+M6/M11). The Definition of Done is met, and now met in a browser:
+
+> The user can create a named Save Point, continue, restart, restore it, and
+> continue differently without losing later history.
+
+```text
+M4 CLOSED — READY FOR M5
+```
+
+M5 is next to brief. It has not been started.
+
+## W.12 What M5 inherits
+
+Recorded here because it is the only thing this report owes the next milestone.
+These are constraints, not suggestions, and none of them is M5's to revisit.
+
+- **The inherited RPG world-state system is instrumentation, not the target
+  architecture.** 60 tests in this milestone use gold arithmetic to make "the
+  state at this position" a number a test can assert. M5 replaces the machinery
+  underneath and must **move the instrumentation while keeping the assertions** —
+  what they measure is where the story is being read and what state belongs to
+  that position, which is exactly as true after M5.
+- **M5 implements ADR 010's genre-neutral typed narrative-state model.** That is
+  the target; the stat/band/cooldown protocol is what it replaces.
+- **Per-position snapshots, or equivalent fast recovery, must survive the
+  replacement** (`TECHNICAL-DESIGN.md` §10.4). Undo, Redo and Save Point restore
+  all resolve a coordinate and read the state recorded there. If M5 makes state
+  reconstruction proportional to campaign length, it degrades all three at once.
+- **M3/M4 history and Save Point semantics are infrastructure now.** The stored
+  head, the single movement mechanism, the capped lineage, fork-on-first-write,
+  and the Save Point coordinate are settled. M5 changes what state *contains*,
+  not how the story is positioned, and should not add a second restore path.
+- **A Save Point's history is protected** (`STORY-BRANCH-SEMANTICS.md` §19.1).
+  Any state-cleanup or migration work M5 introduces must not acquire a way to
+  delete a checkpoint, or the history one names, as a side effect.
+
+---
+
 *End of report.*