fix: quarantine invalid tool results

nanobot/agent/context_governance.py

@@ -232,8 +232,9 @@ class ContextGovernor:
     def drop_orphan_tool_results(
         messages: list[dict[str, Any]],
     ) -> list[dict[str, Any]]:
-        """Drop tool results that have no matching assistant tool_call earlier in history."""
+        """Drop invalid tool results before history is sent back to providers."""
         declared: set[str] = set()
+        fulfilled: set[str] = set()
         updated: list[dict[str, Any]] | None = None
         for idx, msg in enumerate(messages):
             role = msg.get("role")
@@ -243,10 +244,12 @@ class ContextGovernor:
                         declared.add(str(tc["id"]))
             if role == "tool":
                 tid = msg.get("tool_call_id")
-                if tid and str(tid) not in declared:
+                tid_str = str(tid) if tid else ""
+                if not tid_str or tid_str not in declared or tid_str in fulfilled:
                     if updated is None:
                         updated = [dict(m) for m in messages[:idx]]
                     continue
+                fulfilled.add(tid_str)
             if updated is not None:
                 updated.append(dict(msg))
 

nanobot/agent/loop.py

@@ -1753,6 +1753,11 @@ class AgentLoop:
             for tc in m.get("tool_calls") or []
             if isinstance(tc, dict) and tc.get("id")
         }
+        fulfilled_tool_call_ids = {
+            str(m["tool_call_id"])
+            for m in session.messages
+            if m.get("role") == "tool" and m.get("tool_call_id")
+        }
         last_assistant_idx: int | None = None
         for m in messages[skip:]:
             entry = dict(m)
@@ -1767,14 +1772,20 @@ class AgentLoop:
                 continue  # skip empty assistant messages — they poison session context
             if role == "tool":
                 tool_call_id = entry.get("tool_call_id")
-                if not tool_call_id or str(tool_call_id) not in declared_tool_call_ids:
+                tool_call_id_str = str(tool_call_id) if tool_call_id else ""
+                if (
+                    not tool_call_id_str
+                    or tool_call_id_str not in declared_tool_call_ids
+                    or tool_call_id_str in fulfilled_tool_call_ids
+                ):
                     # Undeclared tool results corrupt future provider requests.
                     logger.warning(
-                        "Dropping orphaned tool result {} from session {} during persistence",
+                        "Dropping invalid tool result {} from session {} during persistence",
-                        tool_call_id or "(missing id)",
+                        tool_call_id_str or "(missing id)",
                         session.key,
                     )
                     continue
+                fulfilled_tool_call_ids.add(tool_call_id_str)
                 if isinstance(content, str) and len(content) > self.max_tool_result_chars:
                     entry["content"] = truncate_text_fn(content, self.max_tool_result_chars)
                 elif isinstance(content, list):

tests/agent/test_context_governance.py

@@ -0,0 +1,39 @@
+from nanobot.agent.context_governance import ContextGovernor
+
+
+def _assistant_tool_call(call_id: str) -> dict:
+    return {
+        "role": "assistant",
+        "content": "",
+        "tool_calls": [{
+            "id": call_id,
+            "type": "function",
+            "function": {"name": "exec", "arguments": "{}"},
+        }],
+    }
+
+
+def test_drop_orphan_tool_results_drops_missing_tool_call_id() -> None:
+    messages = [
+        _assistant_tool_call("call_1"),
+        {"role": "tool", "name": "exec", "content": "missing id"},
+        {"role": "tool", "tool_call_id": "call_1", "name": "exec", "content": "ok"},
+    ]
+
+    result = ContextGovernor.drop_orphan_tool_results(messages)
+
+    assert [m.get("tool_call_id") for m in result if m.get("role") == "tool"] == ["call_1"]
+
+
+def test_drop_orphan_tool_results_drops_duplicate_tool_result() -> None:
+    messages = [
+        _assistant_tool_call("call_1"),
+        {"role": "tool", "tool_call_id": "call_1", "name": "exec", "content": "first"},
+        {"role": "tool", "tool_call_id": "call_1", "name": "exec", "content": "duplicate"},
+    ]
+
+    result = ContextGovernor.drop_orphan_tool_results(messages)
+
+    tool_results = [m for m in result if m.get("role") == "tool"]
+    assert len(tool_results) == 1
+    assert tool_results[0]["content"] == "first"

tests/agent/test_loop_save_turn.py

@@ -1870,3 +1870,58 @@ def test_save_turn_keeps_tool_results_declared_in_prior_history() -> None:
     )
 
     assert [m["role"] for m in session.messages] == ["assistant", "tool"]
+
+
+def test_save_turn_drops_tool_result_already_fulfilled_in_history() -> None:
+    loop = _mk_loop()
+    session = Session(key="test:prior-fulfilled")
+    session.add_message(
+        "assistant",
+        "",
+        tool_calls=[{
+            "id": "call_prior",
+            "type": "function",
+            "function": {"name": "exec", "arguments": "{}"},
+        }],
+    )
+    session.add_message(
+        "tool",
+        "first",
+        tool_call_id="call_prior",
+        name="exec",
+    )
+
+    loop._save_turn(
+        session,
+        [{"role": "tool", "tool_call_id": "call_prior", "name": "exec", "content": "duplicate"}],
+        skip=0,
+    )
+
+    assert [m["role"] for m in session.messages] == ["assistant", "tool"]
+    assert session.messages[1]["content"] == "first"
+
+
+def test_save_turn_drops_duplicate_tool_result_ids() -> None:
+    loop = _mk_loop()
+    session = Session(key="test:duplicate-tool-result")
+
+    loop._save_turn(
+        session,
+        [
+            {
+                "role": "assistant",
+                "content": "",
+                "tool_calls": [{
+                    "id": "call_dupe",
+                    "type": "function",
+                    "function": {"name": "exec", "arguments": "{}"},
+                }],
+            },
+            {"role": "tool", "tool_call_id": "call_dupe", "name": "exec", "content": "first"},
+            {"role": "tool", "tool_call_id": "call_dupe", "name": "exec", "content": "second"},
+        ],
+        skip=0,
+    )
+
+    assert [m["role"] for m in session.messages] == ["assistant", "tool"]
+    assert session.messages[1]["content"] == "first"