Fix the run-breaking tool-result order; e2e test runs the slice (M2.1, PR D)

The first live run of the new loop's e2e test found that a fact recorded
while start_milestone was in flight landed between the tool call and its
result; both API formats reject that, so every later request in the run
failed. Views now keep each result right after its call. Scripted test
models check the APIs' ordering rules, so this can't pass unit tests
again.

The e2e tests no longer depend on the sandbox's history: each run asks for
a uniquely named function and works against a base branch of its own,
deleted afterwards. The review step asks for a change that can't already
be there. The scripted operator gives up after five fruitless answers, and
each run keeps its state store for inspection. The driver's prompt says
review fixes stay in the current milestone.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE

tests/integration/test_end_to_end.py

@@ -25,6 +25,8 @@ from pathlib import Path
 from typing import Any
 
 import pytest
+from e2e_support import base_branch, operator_gitea
+from e2e_support import task as make_task
 
 from nisse_dev.agent import Agent, build_agent
 from nisse_dev.canonical import UserMessage
@@ -43,10 +45,6 @@ pytestmark = [pytest.mark.integration, pytest.mark.e2e]
 
 OPERATOR = "@dongar:dongar.net"  # an allowlisted operator in nisse-test.yaml
 SETTLE_S = 1200  # a milestone, with real models, inside one message
-TASK = (
-    "Add a function `count_vowels(text)` to the sandbox package that returns "
-    "how many of a, e, i, o, u (either case) the text contains, with tests."
-)
 
 
 def build_host(ssh_host: SSHHost) -> BuildHost:
@@ -70,10 +68,14 @@ def agent(
     ssh_host: SSHHost,
     tmp_path: Path,
     monkeypatch: pytest.MonkeyPatch,
-) -> Iterator[tuple[Agent, InMemoryTransport]]:
+) -> Iterator[tuple[Agent, InMemoryTransport, str]]:
     run = uuid.uuid4().hex[:8]
+    stack = contextlib.ExitStack()
+    operator = operator_gitea(live_config.gitea_url)
+    base = stack.enter_context(base_branch(operator, live_config.gitea_repo, run))
     config = replace(
         live_config,
+        base_branch=base,
         state_db=tmp_path / "state.db",
         workspace=Path(f"/tmp/nisse-e2e-{run}/sandbox"),
         build_host=build_host(ssh_host),
@@ -84,7 +86,7 @@ def agent(
     try:
         agent = build_agent(config)
         agent.start()
-        yield agent, transport
+        yield agent, transport, run
     finally:
         if agent is not None:
             agent.stop()
@@ -97,6 +99,7 @@ def agent(
             state_dir=tmp_path / "ssh",
         )
         host.execute(f"rm -rf /tmp/nisse-e2e-{run}")
+        stack.close()  # the run's base branch
 
 
 class Operator:
@@ -148,14 +151,16 @@ def is_artifact(path: str) -> bool:
     )
 
 
-def test_a_milestone_end_to_end(agent: tuple[Agent, InMemoryTransport]) -> None:
-    running, transport = agent
+def test_a_milestone_end_to_end(
+    agent: tuple[Agent, InMemoryTransport, str],
+) -> None:
+    running, transport, run = agent
     operator = Operator(running, transport)
     gitea, repo = running.gitea, running.config.gitea_repo
     pr: dict[str, Any] = {}
     try:
         # The task first, before any project: it must be kept, not lost.
-        operator.say(TASK)
+        operator.say(make_task(run))
         operator.say(f"use {running.config.project}")
         state = operator.drive_to_pr()
         pr = state.data["pr"]