fix(platform-integrations): guard non-dict chat JSON in Bob session-id + provenance locators

illeatmyhat · claude · illeatmyhat · commit 8c0066f39d46 · 2026-06-11T10:40:54.000-07:00
A valid-but-non-dict Bob chat file ([], null, scalar) made .get("sessionId")
raise an uncaught AttributeError, crashing the whole audit run (audit_recall.py
_bob_session_id, no uuid fallback) and the entire `provenance candidates` run
(provenance.py bob step-4). Mirror the isinstance(data, dict) guard the codex
branch already uses. Add regression tests seeding a non-dict chat file.

Co-Authored-By: Claude Opus 4.8 (1M context) &lt;noreply@anthropic.com&gt;
diff --git a/platform-integrations/bob/evolve-lite/lib/evolve-lite/audit_recall.py b/platform-integrations/bob/evolve-lite/lib/evolve-lite/audit_recall.py
@@ -57,9 +57,10 @@ def _bob_session_id() -> str | None:
             chats.extend((Path.home() / ".bob" / "tmp" / project_hash / "chats").glob("session-*.json"))
         for chat in sorted(chats, key=lambda p: p.stat().st_mtime, reverse=True):
             try:
-                sid = json.loads(chat.read_text(encoding="utf-8")).get("sessionId")
+                data = json.loads(chat.read_text(encoding="utf-8"))
             except (OSError, json.JSONDecodeError):
                 continue
+            sid = data.get("sessionId") if isinstance(data, dict) else None
             if sid:
                 return str(sid)
     except OSError:
diff --git a/platform-integrations/bob/evolve-lite/skills/evolve-lite-provenance/scripts/provenance.py b/platform-integrations/bob/evolve-lite/skills/evolve-lite-provenance/scripts/provenance.py
@@ -177,7 +177,7 @@ def locate_trajectory(session_id, evolve_dir, *, project_root=None, home=None):
                     data = json.loads(chat.read_text(encoding="utf-8"))
                 except (OSError, json.JSONDecodeError):
                     continue
-                if data.get("sessionId") == session_id:
+                if isinstance(data, dict) and data.get("sessionId") == session_id:
                     return chat
 
     return None
diff --git a/platform-integrations/claude/plugins/evolve-lite/lib/evolve-lite/audit_recall.py b/platform-integrations/claude/plugins/evolve-lite/lib/evolve-lite/audit_recall.py
@@ -57,9 +57,10 @@ def _bob_session_id() -> str | None:
             chats.extend((Path.home() / ".bob" / "tmp" / project_hash / "chats").glob("session-*.json"))
         for chat in sorted(chats, key=lambda p: p.stat().st_mtime, reverse=True):
             try:
-                sid = json.loads(chat.read_text(encoding="utf-8")).get("sessionId")
+                data = json.loads(chat.read_text(encoding="utf-8"))
             except (OSError, json.JSONDecodeError):
                 continue
+            sid = data.get("sessionId") if isinstance(data, dict) else None
             if sid:
                 return str(sid)
     except OSError:
diff --git a/platform-integrations/claude/plugins/evolve-lite/skills/evolve-lite/provenance/scripts/provenance.py b/platform-integrations/claude/plugins/evolve-lite/skills/evolve-lite/provenance/scripts/provenance.py
@@ -177,7 +177,7 @@ def locate_trajectory(session_id, evolve_dir, *, project_root=None, home=None):
                     data = json.loads(chat.read_text(encoding="utf-8"))
                 except (OSError, json.JSONDecodeError):
                     continue
-                if data.get("sessionId") == session_id:
+                if isinstance(data, dict) and data.get("sessionId") == session_id:
                     return chat
 
     return None
diff --git a/platform-integrations/claw-code/plugins/evolve-lite/lib/evolve-lite/audit_recall.py b/platform-integrations/claw-code/plugins/evolve-lite/lib/evolve-lite/audit_recall.py
@@ -57,9 +57,10 @@ def _bob_session_id() -> str | None:
             chats.extend((Path.home() / ".bob" / "tmp" / project_hash / "chats").glob("session-*.json"))
         for chat in sorted(chats, key=lambda p: p.stat().st_mtime, reverse=True):
             try:
-                sid = json.loads(chat.read_text(encoding="utf-8")).get("sessionId")
+                data = json.loads(chat.read_text(encoding="utf-8"))
             except (OSError, json.JSONDecodeError):
                 continue
+            sid = data.get("sessionId") if isinstance(data, dict) else None
             if sid:
                 return str(sid)
     except OSError:
diff --git a/platform-integrations/claw-code/plugins/evolve-lite/skills/evolve-lite/provenance/scripts/provenance.py b/platform-integrations/claw-code/plugins/evolve-lite/skills/evolve-lite/provenance/scripts/provenance.py
@@ -177,7 +177,7 @@ def locate_trajectory(session_id, evolve_dir, *, project_root=None, home=None):
                     data = json.loads(chat.read_text(encoding="utf-8"))
                 except (OSError, json.JSONDecodeError):
                     continue
-                if data.get("sessionId") == session_id:
+                if isinstance(data, dict) and data.get("sessionId") == session_id:
                     return chat
 
     return None
diff --git a/platform-integrations/codex/plugins/evolve-lite/lib/evolve-lite/audit_recall.py b/platform-integrations/codex/plugins/evolve-lite/lib/evolve-lite/audit_recall.py
@@ -57,9 +57,10 @@ def _bob_session_id() -> str | None:
             chats.extend((Path.home() / ".bob" / "tmp" / project_hash / "chats").glob("session-*.json"))
         for chat in sorted(chats, key=lambda p: p.stat().st_mtime, reverse=True):
             try:
-                sid = json.loads(chat.read_text(encoding="utf-8")).get("sessionId")
+                data = json.loads(chat.read_text(encoding="utf-8"))
             except (OSError, json.JSONDecodeError):
                 continue
+            sid = data.get("sessionId") if isinstance(data, dict) else None
             if sid:
                 return str(sid)
     except OSError:
diff --git a/platform-integrations/codex/plugins/evolve-lite/skills/evolve-lite/provenance/scripts/provenance.py b/platform-integrations/codex/plugins/evolve-lite/skills/evolve-lite/provenance/scripts/provenance.py
@@ -177,7 +177,7 @@ def locate_trajectory(session_id, evolve_dir, *, project_root=None, home=None):
                     data = json.loads(chat.read_text(encoding="utf-8"))
                 except (OSError, json.JSONDecodeError):
                     continue
-                if data.get("sessionId") == session_id:
+                if isinstance(data, dict) and data.get("sessionId") == session_id:
                     return chat
 
     return None
diff --git a/plugin-source/lib/audit_recall.py b/plugin-source/lib/audit_recall.py
@@ -57,9 +57,10 @@ def _bob_session_id() -> str | None:
             chats.extend((Path.home() / ".bob" / "tmp" / project_hash / "chats").glob("session-*.json"))
         for chat in sorted(chats, key=lambda p: p.stat().st_mtime, reverse=True):
             try:
-                sid = json.loads(chat.read_text(encoding="utf-8")).get("sessionId")
+                data = json.loads(chat.read_text(encoding="utf-8"))
             except (OSError, json.JSONDecodeError):
                 continue
+            sid = data.get("sessionId") if isinstance(data, dict) else None
             if sid:
                 return str(sid)
     except OSError:
diff --git a/plugin-source/skills/evolve-lite/provenance/scripts/provenance.py b/plugin-source/skills/evolve-lite/provenance/scripts/provenance.py
@@ -177,7 +177,7 @@ def locate_trajectory(session_id, evolve_dir, *, project_root=None, home=None):
                     data = json.loads(chat.read_text(encoding="utf-8"))
                 except (OSError, json.JSONDecodeError):
                     continue
-                if data.get("sessionId") == session_id:
+                if isinstance(data, dict) and data.get("sessionId") == session_id:
                     return chat
 
     return None
diff --git a/tests/platform_integrations/test_audit_recall.py b/tests/platform_integrations/test_audit_recall.py
@@ -150,6 +150,56 @@ def test_bob_picks_newest_chat(self, tmp_path):
         assert rows[0]["session_id"] == newest
         assert "evolve-session:" not in result.stdout
 
+    def test_bob_non_dict_chat_does_not_crash(self, tmp_path):
+        """A valid-but-non-dict newest chat file ([], null, a scalar) must not
+        crash the run: ``.get`` is only called on a dict. With no usable dict
+        chat, the script falls back to a minted uuid (exit 0, row written). When
+        a real dict chat also exists, its sessionId still wins over the non-dict
+        newest file."""
+        import hashlib
+
+        home = tmp_path / "home"
+        proj = tmp_path / "proj"
+        proj.mkdir()
+        # A real dict chat (older).
+        self._seed_bob_chat(home, proj, "dict-sid-1234", filename="session-2026-06-10T20-00-dictsid1.json")
+        # The newest file is valid JSON but NOT a dict — would previously crash.
+        project_hash = hashlib.sha256(os.path.realpath(str(proj)).encode()).hexdigest()
+        bad = home / ".bob" / "tmp" / project_hash / "chats" / "session-2026-06-10T21-30-baadbaad.json"
+        bad.write_text("[]", encoding="utf-8")
+        os.utime(bad, (10**10, 10**10))  # newest => examined first
+
+        result = _run(proj, ["project/baz"], {"BOBSHELL_CLI": "1", "HOME": str(home), "USERPROFILE": str(home)})
+
+        # Did not crash; the non-dict newest is skipped and the real dict chat wins.
+        assert result.returncode == 0
+        rows = _read_rows(proj / ".evolve" / "audit.log")
+        assert len(rows) == 1
+        assert rows[0]["session_id"] == "dict-sid-1234"
+        assert "evolve-session:" not in result.stdout
+
+    def test_bob_non_dict_chat_only_mints_uuid(self, tmp_path):
+        """When the ONLY Bob chat present is valid-but-non-dict JSON (e.g. null),
+        the lookup yields nothing and the script mints a uuid (exit 0, echoed),
+        rather than crashing on ``.get``."""
+        import hashlib
+
+        home = tmp_path / "home"
+        proj = tmp_path / "proj"
+        proj.mkdir()
+        project_hash = hashlib.sha256(os.path.realpath(str(proj)).encode()).hexdigest()
+        chats = home / ".bob" / "tmp" / project_hash / "chats"
+        chats.mkdir(parents=True)
+        (chats / "session-2026-06-10T21-30-baadbaad.json").write_text("null", encoding="utf-8")
+
+        result = _run(proj, ["project/baz"], {"BOBSHELL_CLI": "1", "HOME": str(home), "USERPROFILE": str(home)})
+
+        assert result.returncode == 0
+        rows = _read_rows(proj / ".evolve" / "audit.log")
+        assert len(rows) == 1
+        minted = rows[0]["session_id"]
+        assert f"evolve-session: {minted}" in result.stdout  # minted => echoed
+
     def test_bob_branch_inert_without_bobshell_cli(self, tmp_path):
         """No BOBSHELL_CLI => the Bob lookup never runs (even with a chat present),
         so the script mints a uuid as before. Keeps the branch inert off-Bob."""
diff --git a/tests/platform_integrations/test_provenance.py b/tests/platform_integrations/test_provenance.py
@@ -187,6 +187,30 @@ def test_bob_sessionid_body_mismatch_not_matched(self, tmp_path):
         assert candidates[0]["trajectory_path"] is None
         assert candidates[0]["missing"] == ["trajectory"]
 
+    def test_bob_non_dict_chat_does_not_crash(self, tmp_path):
+        """A chat whose filename prefix matches the sid_head prefilter but whose
+        JSON is valid-but-NON-dict ([], null, a scalar) must not crash the run:
+        ``.get`` is only called on a dict. It simply falls through (no match),
+        so the trajectory is reported missing rather than raising."""
+        home = tmp_path / "home"
+        evolve_dir = tmp_path / "proj" / ".evolve"
+        evolve_dir.mkdir(parents=True)
+        sid = "d6484b2c-24f4-474c-8f43-36544e2dbcd8"
+        write_audit(evolve_dir, [{"event": "recall", "session_id": sid, "entities": ["project/baz"]}])
+        write_entity(evolve_dir, "project/baz")
+        # Filename prefix d6484b2c matches the sid_head prefilter, but the body
+        # is a JSON array (non-dict) — previously crashed on .get("sessionId").
+        chat = home / ".bob" / "tmp" / "abc123hash" / "chats" / "session-2026-06-10T21-12-d6484b2c.json"
+        chat.parent.mkdir(parents=True)
+        chat.write_text("[]", encoding="utf-8")
+
+        result = run_provenance("candidates", evolve_dir=evolve_dir, home=home)
+        assert result.returncode == 0, result.stderr
+        candidates = parse_jsonl(result.stdout)
+        assert len(candidates) == 1
+        assert candidates[0]["trajectory_path"] is None
+        assert candidates[0]["missing"] == ["trajectory"]
+
 
 class TestCandidatesMissing:
     def test_missing_trajectory_still_emitted(self, tmp_path):
