M4.1: toolchain preflight

A model works out and sets up what a project needs on its build VM;
code makes sure that happens, bounds it and checks the result (D3).

- Before each driver run on a project, one round trip checks the build
  VM: the harness's own baseline (missing: the item waits, a template
  problem), the workspace clone (missing: cloned again and said; the
  item resumes from what it pushed) and a per-project marker.
- No marker (a new project, a new or wiped VM): the item's driver model
  runs a bounded preflight first, with its own prompt, the execution
  tools (screened as ever) and toolchain_ready. The harness runs each
  reported version command itself (version-shaped commands only, even
  from the marker, which the model's commands could edit) and writes
  the marker. Gaps and an unfinished preflight are facts and work goes
  on; the thread hears only of installs and gaps. A rebuilt VM starts
  from the last toolchain.
- Later runs check the versions again; a change is a fact and a thread
  line. The PR body says what the work was tested with.
- A changed host key is its own failure (HostKeyChanged): the item
  waits for the operator's `trust new host key`, which drops the pin;
  then the VM is set up again.
- A prompt line: install what the work needs on the build VM, say why,
  and put lasting needs in the project's own pins.
- tuning.preflight: max_turns, command_timeout_s.
- The property walk gains a VM wipe and ready_before_work (checked
  against a mutant that skips preflight); parked_work_safe accepts that
  a wipe loses what wasn't pushed (D3).

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

nisse-dev/tests/test_execution.py

@@ -75,3 +75,34 @@ def test_secrets_are_never_on_a_command_line(ssh_host: SSHHost, tmp_path: Path)
 def test_redact_masks_longest_values_first() -> None:
     assert redact("abc-abcdef", ["abcdef", "abc-abcdef"]) == REDACTED
     assert redact("short ab stays", ["ab"]) == "short ab stays"  # too short to mask
+
+
+def test_a_changed_host_key_is_its_own_failure_and_only_forgetting_clears_it(
+    ssh_host: SSHHost, tmp_path: Path
+) -> None:
+    import subprocess
+
+    from nisse_dev.execution import HostKeyChanged
+
+    host = SSHHost(
+        ssh_host.address,
+        key=ssh_host._key,
+        user=ssh_host.user,
+        port=ssh_host.port,
+        state_dir=tmp_path / "state",
+    )
+    assert host.execute("true").ok  # first contact pins the key
+    host.close()
+    # Someone else's key where the pinned one was: a rebuilt VM, or not it.
+    subprocess.run(
+        ["ssh-keygen", "-q", "-t", "ed25519", "-N", "", "-f", str(tmp_path / "other")],
+        check=True,
+    )
+    other = (tmp_path / "other.pub").read_text().split()
+    known = tmp_path / "state" / "known_hosts"
+    name = host.address if host.port == 22 else f"[{host.address}]:{host.port}"
+    known.write_text(f"{name} {other[0]} {other[1]}\n")
+    with pytest.raises(HostKeyChanged):
+        host.execute("true")
+    host.forget_host_key()
+    assert host.execute("true").ok  # the real key pinned again