feat(mcp): Steward's own backups: the runs, the next one, checking and making one (#5836)

list_backups, check_backup and make_backup are Settings → Backups for an
agent, admin only like the page. make_backup runs the page's job in the
background and is audited as "<user> via MCP". Restoring stays in the
browser. get_overview names a scheduled backup that failed, is two hours
past its slot, or has no bucket.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

tests/integration/test_mcp_backups.py

@@ -0,0 +1,113 @@
+"""Integration (#5836): the MCP's backup tools against real Postgres.
+
+The history Steward keeps in its settings, read back through list_backups;
+make_backup refused with no bucket, then started (the run itself stubbed, as
+in test_backup_job.py) and audited; the overview names a failing backup.
+"""
+from __future__ import annotations
+
+import asyncio
+import os
+import uuid
+from datetime import datetime, timedelta, timezone
+
+import pytest
+
+pytestmark = pytest.mark.integration
+
+
+@pytest.fixture
+def app():
+    if not os.environ.get("STEWARD_DATABASE_URL"):
+        pytest.skip("needs Postgres")
+    from steward.app import create_app
+    return create_app(testing=False)
+
+
+async def _configure(app, **over):
+    from steward.core.settings import set_setting
+    values = {"backup.enabled": True, "backup.s3_endpoint": "http://user:pw@archives:8081",
+              "backup.s3_bucket": "steward", "backup.s3_access_key": "AK",
+              "backup.s3_secret_key": "SK", "backup.schedule_hour": 3, "backup.schedule_tz": "UTC",
+              "backup.history": [], "backup.last_success_at": "", "backup.last_attempt_at": "", **over}
+    async with app.db_sessionmaker() as db:
+        async with db.begin():
+            for k, v in values.items():
+                await set_setting(db, k, v)
+
+
+def test_backups_read_made_and_reported_through_mcp(app, monkeypatch):
+    from sqlalchemy import select
+    from steward.backup import job
+    from steward.core.api_tokens import Caller
+    from steward.mcp import _context
+    from steward.mcp.tools import backups, overview
+    from steward.models.audit import AuditEvent
+    from steward.models.users import User, UserRole
+
+    tag = uuid.uuid4().hex[:8]
+    now = datetime.now(timezone.utc)
+    good_at, failed_at = now - timedelta(days=1), now - timedelta(minutes=10)
+    history = [
+        {"at": failed_at.isoformat(), "ok": False, "by": "schedule", "name": "", "size": 0,
+         "secs": 1.0, "error": "Upload of steward-x.tar refused: 503"},
+        {"at": good_at.isoformat(), "ok": True, "by": f"bk-{tag}", "name": "steward-20261014T030000Z.tar",
+         "size": 52000, "secs": 4.2, "error": "",
+         "check": {"ok": True, "changes": 0, "rows": 310, "summary": "Matches what is in place."}},
+    ]
+    made = []
+
+    async def _fake_run(app_, by="schedule", user_id=None):
+        made.append((by, user_id))
+        return {}
+
+    monkeypatch.setattr(job, "run_backup", _fake_run)
+
+    async def _go():
+        async with app.db_sessionmaker() as s:
+            async with s.begin():
+                user = User(username=f"bk-{tag}", email=f"bk-{tag}@test", password_hash="x",
+                            role=UserRole.admin)
+                s.add(user)
+                await s.flush()
+                user_id = user.id
+        _context._app_ctx.set(app)
+        _context._caller_ctx.set(Caller(user_id=user_id, username=f"bk-{tag}",
+                                        role=UserRole.admin, token_id=9))
+
+        await _configure(app, **{"backup.s3_bucket": ""})
+        with pytest.raises(ValueError, match="bucket isn't set up.*Nothing was started"):
+            await backups.make_backup()
+
+        await _configure(app, **{"backup.history": history,
+                                 "backup.last_success_at": good_at.isoformat(),
+                                 "backup.last_attempt_at": failed_at.isoformat()})
+        listed = await backups.list_backups()
+        assert listed["enabled"] is True and listed["running"] is False
+        assert listed["bucket"] == {"endpoint": "http://user:***@archives:8081", "bucket": "steward",
+                                    "region": "us-east-1"}
+        assert listed["verdict"]["tone"] == "bad" and "503" in listed["verdict"]["error"]
+        assert listed["schedule"]["next_at"] is not None
+        failed, good = listed["runs"]
+        assert failed["ok"] is False and failed["by"] == "schedule" and failed["name"] is None
+        assert good["name"] == "steward-20261014T030000Z.tar" and good["size_bytes"] == 52000
+        assert good["check"]["rows"] == 310 and listed["last_success_age_s"] >= 86000
+        assert (await backups.list_backups(limit=1))["truncated"] is True
+
+        started = await backups.make_backup()
+        assert started["started"] is True
+        await asyncio.sleep(0)          # the background task runs the stub
+        assert made == [(f"bk-{tag} via MCP", user_id)]
+
+        async with app.db_sessionmaker() as s:
+            rows = list((await s.execute(select(AuditEvent).where(
+                AuditEvent.user_id == user_id, AuditEvent.action == "backup.started"))).scalars())
+        assert len(rows) == 1 and rows[0].username == f"bk-{tag} via MCP"
+
+        ov = await overview.get_overview()
+        mine = [f for f in ov["findings"] if f["kind"] == "backup_failed"]
+        assert len(mine) == 1 and "refused: 503" in mine[0]["summary"]
+
+        await _configure(app, **{"backup.enabled": False})
+
+    asyncio.run(_go())