Answer where asked, wait on a merge, and let the driver title PRs

From the new loop's first live run (sandbox #255, PR #256):
- A question asked on the PR is answered on the PR (an inline one quoted
  with its file and line, since the API can't reply in its thread), with
  a pointer in the item's Matrix thread; one asked in Matrix stays there.
- wait_for_review: when a review needs no change, or only the operator's
  merge is left, the item waits on the merge instead of on an answer.
- finish_milestone takes the PR's title from the driver.
The e2e test's merge waits out Gitea's brief 'try again later' after a
push.

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

nisse-dev/tests/test_loop_milestones.py

@@ -80,7 +80,13 @@ class Forge:
 
     def open_pull_request(self, repo: str, *, head, base, title, body) -> PullRequest:
         n = 100 + len(self.prs)
-        self.prs[n] = {"head": head, "state": "open", "merged": False, "body": body}
+        self.prs[n] = {
+            "head": head,
+            "state": "open",
+            "merged": False,
+            "body": body,
+            "title": title,
+        }
         return self._pr(n)
 
     def pull_request(self, repo: str, index: int) -> PullRequest:
@@ -476,3 +482,89 @@ def test_abandoning_an_item_closes_its_open_pr(world: dict[str, Any]) -> None:
     assert item(world).outcome == "abandoned"
     assert forge.prs[pr]["state"] == "closed"
     assert (pr, "abandoned: abandoned by @don:test") in forge.posted
+
+
+# --- answering where asked (D5), waiting for review, PR titles -----------------
+
+
+def test_a_question_on_the_pr_is_answered_on_the_pr(world: dict[str, Any]) -> None:
+    forge: Forge = world["forge"]
+    model = ScriptedModel(
+        RESTATE,
+        WORK + FINISH,
+        [call("ask_human", question="bool is a subclass of int; keep the check?")],
+    )
+    agent, transport = loop_agent(world, model)
+    agent.intake()
+    say(transport, world, "go")
+    [pr] = forge.prs
+    forge.notes[pr] = [ReviewNote("line-9", "don", "Is bool an int?", "text.py:9")]
+    agent.watch()
+    agent._schedule()
+    number, body = forge.posted[-1]
+    assert number == pr
+    assert body.startswith("> on `text.py:9`:\n> Is bool an int?\n\n")
+    assert body.endswith("keep the check?")
+    # Matrix gets a pointer, not the conversation.
+    assert (
+        transport.outbound[-1].text
+        == f"I answered your comment on PR #{pr}: https://g/pr/{pr}"
+    )
+    assert transport.outbound[-1].thread == loop_of(item(world)).thread
+
+
+def test_after_a_matrix_reply_questions_go_back_to_matrix(
+    world: dict[str, Any],
+) -> None:
+    forge: Forge = world["forge"]
+    model = ScriptedModel(
+        RESTATE,
+        WORK + FINISH,
+        [call("ask_human", question="first?")],
+        [call("ask_human", question="second?")],
+    )
+    agent, transport = loop_agent(world, model)
+    agent.intake()
+    say(transport, world, "go")
+    [pr] = forge.prs
+    forge.comments[pr] = [Comment(3, "don", "why this?", "t")]
+    agent.watch()
+    agent._schedule()
+    posted_on_pr = len(forge.posted)
+    say(transport, world, "answering in Matrix")  # the operator moved
+    assert len(forge.posted) == posted_on_pr
+    assert transport.outbound[-1].text == "second?"
+
+
+def test_wait_for_review_waits_on_the_merge_then_closes(world: dict[str, Any]) -> None:
+    forge: Forge = world["forge"]
+    model = ScriptedModel(
+        RESTATE,
+        WORK + FINISH,
+        [call("wait_for_review", note="no change needed; ready to merge")],
+        [call("close_item", summary="merged")],
+    )
+    agent, transport = loop_agent(world, model)
+    agent.intake()
+    say(transport, world, "go")
+    [pr] = forge.prs
+    forge.comments[pr] = [Comment(4, "don", "looks fine, merge it?", "t")]
+    agent.watch()
+    agent._schedule()
+    loop = loop_of(item(world))
+    assert (loop.wait, loop.note) == (MERGE, f"PR #{pr}")
+    assert forge.posted[-1][1].endswith("no change needed; ready to merge")
+    forge.merge(pr)
+    agent.watch()
+    agent._schedule()
+    assert item(world).status == "closed"
+
+
+def test_the_driver_titles_its_pr(world: dict[str, Any]) -> None:
+    forge: Forge = world["forge"]
+    finish = [call("finish_milestone", summary="s", title="Add count_words")]
+    agent, transport = loop_agent(world, ScriptedModel(RESTATE, WORK + finish))
+    agent.intake()
+    say(transport, world, "go")
+    [pr] = forge.prs
+    assert forge.prs[pr]["title"] == "Add count_words"

nisse-dev/tests/test_loop_properties.py

@@ -53,6 +53,8 @@ class PolicyModel:
             return [ToolCall(call_id=f"p{self.n}", tool="nonsense", arguments={})]
         if "finish_milestone" not in offered:
             return [call("restate", text=f"restatement {self.n}")]
+        if "wait_for_review" in offered and self.rng.random() < 0.2:
+            return [call("wait_for_review", note=f"waiting {self.n}")]
         merged = any("merged" in t for t in facts[-6:])
         if merged and self.rng.random() < 0.8:
             return [call("close_item", summary=f"done {self.n}")]