Merge branch 'main' of github.com:NousResearch/hermes-agent into feat/ink-refactor

tests/gateway/test_auto_continue.py

@@ -0,0 +1,95 @@
+"""Tests for the auto-continue feature (#4493).
+
+When the gateway restarts mid-agent-work, the session transcript ends on a
+tool result that the agent never processed.  The auto-continue logic detects
+this and prepends a system note to the next user message so the model
+finishes the interrupted work before addressing the new input.
+"""
+
+import pytest
+
+
+def _simulate_auto_continue(agent_history: list, user_message: str) -> str:
+    """Reproduce the auto-continue injection logic from _run_agent().
+
+    This mirrors the exact code in gateway/run.py so we can test the
+    detection and message transformation without spinning up a full
+    gateway runner.
+    """
+    message = user_message
+    if agent_history and agent_history[-1].get("role") == "tool":
+        message = (
+            "[System note: Your previous turn was interrupted before you could "
+            "process the last tool result(s). The conversation history contains "
+            "tool outputs you haven't responded to yet. Please finish processing "
+            "those results and summarize what was accomplished, then address the "
+            "user's new message below.]\n\n"
+            + message
+        )
+    return message
+
+
+class TestAutoDetection:
+    """Test that trailing tool results are correctly detected."""
+
+    def test_trailing_tool_result_triggers_note(self):
+        history = [
+            {"role": "user", "content": "deploy the app"},
+            {"role": "assistant", "content": None, "tool_calls": [
+                {"id": "call_1", "function": {"name": "terminal", "arguments": "{}"}}
+            ]},
+            {"role": "tool", "tool_call_id": "call_1", "content": "deployed successfully"},
+        ]
+        result = _simulate_auto_continue(history, "what happened?")
+        assert "[System note:" in result
+        assert "interrupted" in result
+        assert "what happened?" in result
+
+    def test_trailing_assistant_message_no_note(self):
+        history = [
+            {"role": "user", "content": "hello"},
+            {"role": "assistant", "content": "Hi there!"},
+        ]
+        result = _simulate_auto_continue(history, "how are you?")
+        assert "[System note:" not in result
+        assert result == "how are you?"
+
+    def test_empty_history_no_note(self):
+        result = _simulate_auto_continue([], "hello")
+        assert result == "hello"
+
+    def test_trailing_user_message_no_note(self):
+        """Shouldn't happen in practice, but ensure no false positive."""
+        history = [
+            {"role": "user", "content": "hello"},
+        ]
+        result = _simulate_auto_continue(history, "hello again")
+        assert result == "hello again"
+
+    def test_multiple_tool_results_still_triggers(self):
+        """Multiple tool calls in a row — last one is still role=tool."""
+        history = [
+            {"role": "user", "content": "search and read"},
+            {"role": "assistant", "content": None, "tool_calls": [
+                {"id": "call_1", "function": {"name": "search", "arguments": "{}"}},
+                {"id": "call_2", "function": {"name": "read", "arguments": "{}"}},
+            ]},
+            {"role": "tool", "tool_call_id": "call_1", "content": "found it"},
+            {"role": "tool", "tool_call_id": "call_2", "content": "file content here"},
+        ]
+        result = _simulate_auto_continue(history, "continue")
+        assert "[System note:" in result
+
+    def test_original_message_preserved_after_note(self):
+        """The user's actual message must appear after the system note."""
+        history = [
+            {"role": "assistant", "content": None, "tool_calls": [
+                {"id": "c1", "function": {"name": "t", "arguments": "{}"}}
+            ]},
+            {"role": "tool", "tool_call_id": "c1", "content": "done"},
+        ]
+        result = _simulate_auto_continue(history, "now do X")
+        # System note comes first, then user's message
+        note_end = result.index("]\n\n")
+        user_msg_start = result.index("now do X")
+        assert user_msg_start > note_end

tests/gateway/test_discord_slash_commands.py

@@ -19,10 +19,34 @@ def _ensure_discord_mock():
     discord_mod.Thread = type("Thread", (), {})
     discord_mod.ForumChannel = type("ForumChannel", (), {})
     discord_mod.Interaction = object
+
+    # Lightweight mock for app_commands.Group and Command used by
+    # _register_skill_group.
+    class _FakeGroup:
+        def __init__(self, *, name, description, parent=None):
+            self.name = name
+            self.description = description
+            self.parent = parent
+            self._children: dict[str, object] = {}
+            if parent is not None:
+                parent.add_command(self)
+
+        def add_command(self, cmd):
+            self._children[cmd.name] = cmd
+
+    class _FakeCommand:
+        def __init__(self, *, name, description, callback, parent=None):
+            self.name = name
+            self.description = description
+            self.callback = callback
+            self.parent = parent
+
     discord_mod.app_commands = SimpleNamespace(
         describe=lambda **kwargs: (lambda fn: fn),
         choices=lambda **kwargs: (lambda fn: fn),
         Choice=lambda **kwargs: SimpleNamespace(**kwargs),
+        Group=_FakeGroup,
+        Command=_FakeCommand,
     )
 
     ext_mod = MagicMock()
@@ -51,6 +75,12 @@ class FakeTree:
 
         return decorator
 
+    def add_command(self, cmd):
+        self.commands[cmd.name] = cmd
+
+    def get_commands(self):
+        return [SimpleNamespace(name=n) for n in self.commands]
+
 
 @pytest.fixture
 def adapter():
@@ -87,6 +117,23 @@ async def test_registers_native_thread_slash_command(adapter):
     adapter._handle_thread_create_slash.assert_awaited_once_with(interaction, "Planning", "", 1440)
 
 
+@pytest.mark.asyncio
+async def test_registers_native_restart_slash_command(adapter):
+    adapter._run_simple_slash = AsyncMock()
+    adapter._register_slash_commands()
+
+    assert "restart" in adapter._client.tree.commands
+
+    interaction = SimpleNamespace()
+    await adapter._client.tree.commands["restart"](interaction)
+
+    adapter._run_simple_slash.assert_awaited_once_with(
+        interaction,
+        "/restart",
+        "Restart requested~",
+    )
+
+
 # ------------------------------------------------------------------
 # _handle_thread_create_slash — success, session dispatch, failure
 # ------------------------------------------------------------------
@@ -498,3 +545,79 @@ def test_discord_auto_thread_config_bridge(monkeypatch, tmp_path):
 
     import os
     assert os.getenv("DISCORD_AUTO_THREAD") == "true"
+
+
+# ------------------------------------------------------------------
+# /skill group registration
+# ------------------------------------------------------------------
+
+
+def test_register_skill_group_creates_group(adapter):
+    """_register_skill_group should register a '/skill' Group on the tree."""
+    mock_categories = {
+        "creative": [
+            ("ascii-art", "Generate ASCII art", "/ascii-art"),
+            ("excalidraw", "Hand-drawn diagrams", "/excalidraw"),
+        ],
+        "media": [
+            ("gif-search", "Search for GIFs", "/gif-search"),
+        ],
+    }
+    mock_uncategorized = [
+        ("dogfood", "Exploratory QA testing", "/dogfood"),
+    ]
+
+    with patch(
+        "hermes_cli.commands.discord_skill_commands_by_category",
+        return_value=(mock_categories, mock_uncategorized, 0),
+    ):
+        adapter._register_slash_commands()
+
+    tree = adapter._client.tree
+    assert "skill" in tree.commands, "Expected /skill group to be registered"
+    skill_group = tree.commands["skill"]
+    assert skill_group.name == "skill"
+    # Should have 2 category subgroups + 1 uncategorized subcommand
+    children = skill_group._children
+    assert "creative" in children
+    assert "media" in children
+    assert "dogfood" in children
+    # Category groups should have their skills
+    assert "ascii-art" in children["creative"]._children
+    assert "excalidraw" in children["creative"]._children
+    assert "gif-search" in children["media"]._children
+
+
+def test_register_skill_group_empty_skills_no_group(adapter):
+    """No /skill group should be added when there are zero skills."""
+    with patch(
+        "hermes_cli.commands.discord_skill_commands_by_category",
+        return_value=({}, [], 0),
+    ):
+        adapter._register_slash_commands()
+
+    tree = adapter._client.tree
+    assert "skill" not in tree.commands
+
+
+def test_register_skill_group_handler_dispatches_command(adapter):
+    """Skill subcommand handlers should dispatch the correct /cmd-key text."""
+    mock_categories = {
+        "media": [
+            ("gif-search", "Search for GIFs", "/gif-search"),
+        ],
+    }
+
+    with patch(
+        "hermes_cli.commands.discord_skill_commands_by_category",
+        return_value=(mock_categories, [], 0),
+    ):
+        adapter._register_slash_commands()
+
+    skill_group = adapter._client.tree.commands["skill"]
+    media_group = skill_group._children["media"]
+    gif_cmd = media_group._children["gif-search"]
+    assert gif_cmd.callback is not None
+    # The callback name should reflect the skill
+    assert "gif_search" in gif_cmd.callback.__name__
+

tests/gateway/test_restart_drain.py

@@ -193,7 +193,7 @@ async def test_shutdown_notification_says_restarting_when_restart_requested():
 
     assert len(adapter.sent) == 1
     assert "restarting" in adapter.sent[0]
-    assert "/retry" in adapter.sent[0]
+    assert "resume" in adapter.sent[0]
 
 
 @pytest.mark.asyncio

tests/gateway/test_stuck_loop.py

@@ -0,0 +1,116 @@
+"""Tests for stuck-session loop detection (#7536).
+
+When a session is active across 3+ consecutive gateway restarts (the agent
+gets stuck, gateway restarts, same session gets stuck again), the session
+is auto-suspended on startup so the user gets a clean slate.
+"""
+
+import json
+from pathlib import Path
+from unittest.mock import MagicMock
+
+import pytest
+
+from tests.gateway.restart_test_helpers import make_restart_runner
+
+
+@pytest.fixture
+def runner_with_home(tmp_path, monkeypatch):
+    """Create a runner with a writable HERMES_HOME."""
+    monkeypatch.setattr("gateway.run._hermes_home", tmp_path)
+    runner, adapter = make_restart_runner()
+    return runner, tmp_path
+
+
+class TestStuckLoopDetection:
+
+    def test_increment_creates_file(self, runner_with_home):
+        runner, home = runner_with_home
+        runner._increment_restart_failure_counts({"session:a", "session:b"})
+        path = home / runner._STUCK_LOOP_FILE
+        assert path.exists()
+        counts = json.loads(path.read_text())
+        assert counts["session:a"] == 1
+        assert counts["session:b"] == 1
+
+    def test_increment_accumulates(self, runner_with_home):
+        runner, home = runner_with_home
+        runner._increment_restart_failure_counts({"session:a"})
+        runner._increment_restart_failure_counts({"session:a"})
+        runner._increment_restart_failure_counts({"session:a"})
+        counts = json.loads((home / runner._STUCK_LOOP_FILE).read_text())
+        assert counts["session:a"] == 3
+
+    def test_increment_drops_inactive_sessions(self, runner_with_home):
+        runner, home = runner_with_home
+        runner._increment_restart_failure_counts({"session:a", "session:b"})
+        runner._increment_restart_failure_counts({"session:a"})  # b not active
+        counts = json.loads((home / runner._STUCK_LOOP_FILE).read_text())
+        assert "session:a" in counts
+        assert "session:b" not in counts
+
+    def test_suspend_at_threshold(self, runner_with_home):
+        runner, home = runner_with_home
+        # Simulate 3 restarts with session:a active each time
+        for _ in range(3):
+            runner._increment_restart_failure_counts({"session:a"})
+
+        # Create a mock session entry
+        mock_entry = MagicMock()
+        mock_entry.suspended = False
+        runner.session_store._entries = {"session:a": mock_entry}
+        runner.session_store._save = MagicMock()
+
+        suspended = runner._suspend_stuck_loop_sessions()
+        assert suspended == 1
+        assert mock_entry.suspended is True
+
+    def test_no_suspend_below_threshold(self, runner_with_home):
+        runner, home = runner_with_home
+        runner._increment_restart_failure_counts({"session:a"})
+        runner._increment_restart_failure_counts({"session:a"})
+        # Only 2 restarts — below threshold of 3
+
+        mock_entry = MagicMock()
+        mock_entry.suspended = False
+        runner.session_store._entries = {"session:a": mock_entry}
+
+        suspended = runner._suspend_stuck_loop_sessions()
+        assert suspended == 0
+        assert mock_entry.suspended is False
+
+    def test_clear_on_success(self, runner_with_home):
+        runner, home = runner_with_home
+        runner._increment_restart_failure_counts({"session:a", "session:b"})
+        runner._clear_restart_failure_count("session:a")
+
+        path = home / runner._STUCK_LOOP_FILE
+        counts = json.loads(path.read_text())
+        assert "session:a" not in counts
+        assert "session:b" in counts
+
+    def test_clear_removes_file_when_empty(self, runner_with_home):
+        runner, home = runner_with_home
+        runner._increment_restart_failure_counts({"session:a"})
+        runner._clear_restart_failure_count("session:a")
+        assert not (home / runner._STUCK_LOOP_FILE).exists()
+
+    def test_suspend_clears_file(self, runner_with_home):
+        runner, home = runner_with_home
+        for _ in range(3):
+            runner._increment_restart_failure_counts({"session:a"})
+
+        mock_entry = MagicMock()
+        mock_entry.suspended = False
+        runner.session_store._entries = {"session:a": mock_entry}
+        runner.session_store._save = MagicMock()
+
+        runner._suspend_stuck_loop_sessions()
+        assert not (home / runner._STUCK_LOOP_FILE).exists()
+
+    def test_no_file_no_crash(self, runner_with_home):
+        runner, home = runner_with_home
+        # No file exists — should return 0 and not crash
+        assert runner._suspend_stuck_loop_sessions() == 0
+        # Clear on nonexistent file — should not crash
+        runner._clear_restart_failure_count("nonexistent")

tests/hermes_cli/test_commands.py

@@ -1031,3 +1031,154 @@ class TestDiscordSkillCommands:
             assert len(name) <= _CMD_NAME_LIMIT, (
                 f"Name '{name}' is {len(name)} chars (limit {_CMD_NAME_LIMIT})"
             )
+
+
+# ---------------------------------------------------------------------------
+# Discord skill commands grouped by category
+# ---------------------------------------------------------------------------
+
+from hermes_cli.commands import discord_skill_commands_by_category  # noqa: E402
+
+
+class TestDiscordSkillCommandsByCategory:
+    """Tests for discord_skill_commands_by_category() — /skill group registration."""
+
+    def test_groups_skills_by_category(self, tmp_path, monkeypatch):
+        """Skills nested 2+ levels deep should be grouped by top-level category."""
+        from unittest.mock import patch
+
+        fake_skills_dir = str(tmp_path / "skills")
+        # Create the directory structure so resolve() works
+        for p in [
+            "skills/creative/ascii-art",
+            "skills/creative/excalidraw",
+            "skills/media/gif-search",
+        ]:
+            (tmp_path / p).mkdir(parents=True, exist_ok=True)
+            (tmp_path / p / "SKILL.md").write_text("---\nname: test\n---\n")
+
+        fake_cmds = {
+            "/ascii-art": {
+                "name": "ascii-art",
+                "description": "Generate ASCII art",
+                "skill_md_path": f"{fake_skills_dir}/creative/ascii-art/SKILL.md",
+            },
+            "/excalidraw": {
+                "name": "excalidraw",
+                "description": "Hand-drawn diagrams",
+                "skill_md_path": f"{fake_skills_dir}/creative/excalidraw/SKILL.md",
+            },
+            "/gif-search": {
+                "name": "gif-search",
+                "description": "Search for GIFs",
+                "skill_md_path": f"{fake_skills_dir}/media/gif-search/SKILL.md",
+            },
+        }
+        monkeypatch.setenv("HERMES_HOME", str(tmp_path))
+        with (
+            patch("agent.skill_commands.get_skill_commands", return_value=fake_cmds),
+            patch("tools.skills_tool.SKILLS_DIR", tmp_path / "skills"),
+        ):
+            categories, uncategorized, hidden = discord_skill_commands_by_category(
+                reserved_names=set(),
+            )
+
+        assert "creative" in categories
+        assert "media" in categories
+        assert len(categories["creative"]) == 2
+        assert len(categories["media"]) == 1
+        assert uncategorized == []
+        assert hidden == 0
+
+    def test_root_level_skills_are_uncategorized(self, tmp_path, monkeypatch):
+        """Skills directly under SKILLS_DIR (only 1 path component) → uncategorized."""
+        from unittest.mock import patch
+
+        fake_skills_dir = str(tmp_path / "skills")
+        (tmp_path / "skills" / "dogfood").mkdir(parents=True, exist_ok=True)
+        (tmp_path / "skills" / "dogfood" / "SKILL.md").write_text("")
+
+        fake_cmds = {
+            "/dogfood": {
+                "name": "dogfood",
+                "description": "QA testing",
+                "skill_md_path": f"{fake_skills_dir}/dogfood/SKILL.md",
+            },
+        }
+        monkeypatch.setenv("HERMES_HOME", str(tmp_path))
+        with (
+            patch("agent.skill_commands.get_skill_commands", return_value=fake_cmds),
+            patch("tools.skills_tool.SKILLS_DIR", tmp_path / "skills"),
+        ):
+            categories, uncategorized, hidden = discord_skill_commands_by_category(
+                reserved_names=set(),
+            )
+
+        assert categories == {}
+        assert len(uncategorized) == 1
+        assert uncategorized[0][0] == "dogfood"
+
+    def test_hub_skills_excluded(self, tmp_path, monkeypatch):
+        """Skills under .hub should be excluded."""
+        from unittest.mock import patch
+
+        fake_skills_dir = str(tmp_path / "skills")
+        (tmp_path / "skills" / ".hub" / "some-skill").mkdir(parents=True, exist_ok=True)
+        (tmp_path / "skills" / ".hub" / "some-skill" / "SKILL.md").write_text("")
+
+        fake_cmds = {
+            "/some-skill": {
+                "name": "some-skill",
+                "description": "Hub skill",
+                "skill_md_path": f"{fake_skills_dir}/.hub/some-skill/SKILL.md",
+            },
+        }
+        monkeypatch.setenv("HERMES_HOME", str(tmp_path))
+        with (
+            patch("agent.skill_commands.get_skill_commands", return_value=fake_cmds),
+            patch("tools.skills_tool.SKILLS_DIR", tmp_path / "skills"),
+        ):
+            categories, uncategorized, hidden = discord_skill_commands_by_category(
+                reserved_names=set(),
+            )
+
+        assert categories == {}
+        assert uncategorized == []
+
+    def test_deep_nested_skills_use_top_category(self, tmp_path, monkeypatch):
+        """Skills like mlops/training/axolotl should group under 'mlops'."""
+        from unittest.mock import patch
+
+        fake_skills_dir = str(tmp_path / "skills")
+        (tmp_path / "skills" / "mlops" / "training" / "axolotl").mkdir(parents=True, exist_ok=True)
+        (tmp_path / "skills" / "mlops" / "training" / "axolotl" / "SKILL.md").write_text("")
+        (tmp_path / "skills" / "mlops" / "inference" / "vllm").mkdir(parents=True, exist_ok=True)
+        (tmp_path / "skills" / "mlops" / "inference" / "vllm" / "SKILL.md").write_text("")
+
+        fake_cmds = {
+            "/axolotl": {
+                "name": "axolotl",
+                "description": "Fine-tuning with Axolotl",
+                "skill_md_path": f"{fake_skills_dir}/mlops/training/axolotl/SKILL.md",
+            },
+            "/vllm": {
+                "name": "vllm",
+                "description": "vLLM inference",
+                "skill_md_path": f"{fake_skills_dir}/mlops/inference/vllm/SKILL.md",
+            },
+        }
+        monkeypatch.setenv("HERMES_HOME", str(tmp_path))
+        with (
+            patch("agent.skill_commands.get_skill_commands", return_value=fake_cmds),
+            patch("tools.skills_tool.SKILLS_DIR", tmp_path / "skills"),
+        ):
+            categories, uncategorized, hidden = discord_skill_commands_by_category(
+                reserved_names=set(),
+            )
+
+        # Both should be under 'mlops' regardless of sub-category
+        assert "mlops" in categories
+        names = {n for n, _d, _k in categories["mlops"]}
+        assert "axolotl" in names
+        assert "vllm" in names
+        assert len(uncategorized) == 0

tests/hermes_cli/test_gateway_service.py

@@ -452,7 +452,7 @@ class TestGatewayServiceDetection:
 
 
 class TestGatewaySystemServiceRouting:
-    def test_systemd_restart_self_requests_graceful_restart_without_reload_or_restart(self, monkeypatch, capsys):
+    def test_systemd_restart_self_requests_graceful_restart_and_waits(self, monkeypatch, capsys):
         calls = []
 
         monkeypatch.setattr(gateway_cli, "_select_systemd_scope", lambda system=False: False)
@@ -466,16 +466,37 @@ class TestGatewaySystemServiceRouting:
             "_request_gateway_self_restart",
             lambda pid: calls.append(("self", pid)) or True,
         )
-        monkeypatch.setattr(
-            gateway_cli.subprocess,
-            "run",
-            lambda *args, **kwargs: (_ for _ in ()).throw(AssertionError("systemctl should not run")),
-        )
+
+        # Simulate: old process dies immediately, new process becomes active
+        kill_call_count = [0]
+        def fake_kill(pid, sig):
+            kill_call_count[0] += 1
+            if kill_call_count[0] >= 2:  # first call checks, second = dead
+                raise ProcessLookupError()
+        monkeypatch.setattr(os, "kill", fake_kill)
+
+        # Simulate systemctl is-active returning "active" with a new PID
+        new_pid = [None]
+        def fake_subprocess_run(cmd, **kwargs):
+            if "is-active" in cmd:
+                result = SimpleNamespace(stdout="active\n", returncode=0)
+                new_pid[0] = 999  # new PID
+                return result
+            raise AssertionError(f"Unexpected systemctl call: {cmd}")
+
+        monkeypatch.setattr(gateway_cli.subprocess, "run", fake_subprocess_run)
+        # get_running_pid returns new PID after restart
+        pid_calls = [0]
+        def fake_get_pid():
+            pid_calls[0] += 1
+            return 999 if pid_calls[0] > 1 else 654
+        monkeypatch.setattr("gateway.status.get_running_pid", fake_get_pid)
 
         gateway_cli.systemd_restart()
 
-        assert calls == [("refresh", False), ("self", 654)]
-        assert "restart requested" in capsys.readouterr().out.lower()
+        assert ("self", 654) in calls
+        out = capsys.readouterr().out.lower()
+        assert "restarted" in out
 
     def test_gateway_install_passes_system_flags(self, monkeypatch):
         monkeypatch.setattr(gateway_cli, "supports_systemd_services", lambda: True)

tests/test_toolsets.py

@@ -116,6 +116,22 @@ class TestValidateToolset:
     def test_invalid(self):
         assert validate_toolset("nonexistent") is False
 
+    def test_mcp_alias_uses_live_registry(self, monkeypatch):
+        reg = ToolRegistry()
+        reg.register(
+            name="mcp_dynserver_ping",
+            toolset="mcp-dynserver",
+            schema=_make_schema("mcp_dynserver_ping", "Ping"),
+            handler=_dummy_handler,
+        )
+        reg.register_toolset_alias("dynserver", "mcp-dynserver")
+
+        monkeypatch.setattr("tools.registry.registry", reg)
+
+        assert validate_toolset("dynserver") is True
+        assert validate_toolset("mcp-dynserver") is True
+        assert "mcp_dynserver_ping" in resolve_toolset("dynserver")
+
 
 class TestGetToolsetInfo:
     def test_leaf(self):
@@ -150,6 +166,23 @@ class TestCreateCustomToolset:
             del TOOLSETS["_test_custom"]
 
 
+class TestRegistryOwnedToolsets:
+    def test_registry_membership_is_live(self, monkeypatch):
+        reg = ToolRegistry()
+        reg.register(
+            name="test_live_toolset_tool",
+            toolset="test-live-toolset",
+            schema=_make_schema("test_live_toolset_tool", "Live"),
+            handler=_dummy_handler,
+        )
+
+        monkeypatch.setattr("tools.registry.registry", reg)
+
+        assert validate_toolset("test-live-toolset") is True
+        assert get_toolset("test-live-toolset")["tools"] == ["test_live_toolset_tool"]
+        assert resolve_toolset("test-live-toolset") == ["test_live_toolset_tool"]
+
+
 class TestToolsetConsistency:
     """Verify structural integrity of the built-in TOOLSETS dict."""
 

tests/tools/test_browser_homebrew_paths.py

@@ -31,18 +31,25 @@ def _clear_browser_caches():
 
 
 class TestSanePath:
-    """Verify _SANE_PATH includes Homebrew directories."""
+    """Verify _SANE_PATH includes fallback directories used by browser_tool."""
+
+    def test_includes_termux_bin(self):
+        assert "/data/data/com.termux/files/usr/bin" in _SANE_PATH.split(os.pathsep)
+
+    def test_includes_termux_sbin(self):
+        assert "/data/data/com.termux/files/usr/sbin" in _SANE_PATH.split(os.pathsep)
 
     def test_includes_homebrew_bin(self):
-        assert "/opt/homebrew/bin" in _SANE_PATH
+        assert "/opt/homebrew/bin" in _SANE_PATH.split(os.pathsep)
 
     def test_includes_homebrew_sbin(self):
-        assert "/opt/homebrew/sbin" in _SANE_PATH
+        assert "/opt/homebrew/sbin" in _SANE_PATH.split(os.pathsep)
 
     def test_includes_standard_dirs(self):
-        assert "/usr/local/bin" in _SANE_PATH
-        assert "/usr/bin" in _SANE_PATH
-        assert "/bin" in _SANE_PATH
+        path_parts = _SANE_PATH.split(os.pathsep)
+        assert "/usr/local/bin" in path_parts
+        assert "/usr/bin" in path_parts
+        assert "/bin" in path_parts
 
 
 class TestDiscoverHomebrewNodeDirs:
@@ -143,6 +150,44 @@ class TestFindAgentBrowser:
             result = _find_agent_browser()
             assert result == "npx agent-browser"
 
+    def test_finds_npx_in_termux_fallback_path(self):
+        """Should find npx when only Termux fallback dirs are available."""
+        def mock_which(cmd, path=None):
+            if cmd == "agent-browser":
+                return None
+            if cmd == "npx":
+                if path and "/data/data/com.termux/files/usr/bin" in path:
+                    return "/data/data/com.termux/files/usr/bin/npx"
+                return None
+            return None
+
+        original_path_exists = Path.exists
+
+        def mock_path_exists(self):
+            if "node_modules" in str(self) and "agent-browser" in str(self):
+                return False
+            return original_path_exists(self)
+
+        real_isdir = os.path.isdir
+
+        def selective_isdir(path):
+            if path in (
+                "/data/data/com.termux/files/usr/bin",
+                "/data/data/com.termux/files/usr/sbin",
+            ):
+                return True
+            return real_isdir(path)
+
+        with patch("shutil.which", side_effect=mock_which), \
+             patch("os.path.isdir", side_effect=selective_isdir), \
+             patch.object(Path, "exists", mock_path_exists), \
+             patch(
+                 "tools.browser_tool._discover_homebrew_node_dirs",
+                 return_value=[],
+             ):
+            result = _find_agent_browser()
+            assert result == "npx agent-browser"
+
     def test_raises_when_not_found(self):
         """Should raise FileNotFoundError when nothing works."""
         original_path_exists = Path.exists
@@ -399,3 +444,51 @@ class TestRunBrowserCommandPathConstruction:
         result_path = captured_env.get("PATH", "")
         assert "/opt/homebrew/bin" in result_path
         assert "/opt/homebrew/sbin" in result_path
+
+    def test_subprocess_path_includes_termux_fallback_dirs(self, tmp_path):
+        """Termux fallback dirs should survive browser PATH rebuilding."""
+        captured_env = {}
+
+        mock_proc = MagicMock()
+        mock_proc.returncode = 0
+        mock_proc.wait.return_value = 0
+
+        def capture_popen(cmd, **kwargs):
+            captured_env.update(kwargs.get("env", {}))
+            return mock_proc
+
+        fake_session = {
+            "session_name": "test-session",
+            "session_id": "test-id",
+            "cdp_url": None,
+        }
+
+        fake_json = json.dumps({"success": True})
+        real_isdir = os.path.isdir
+
+        def selective_isdir(path):
+            if path in (
+                "/data/data/com.termux/files/usr/bin",
+                "/data/data/com.termux/files/usr/sbin",
+            ):
+                return True
+            if path.startswith(str(tmp_path)):
+                return True
+            return real_isdir(path)
+
+        with patch("tools.browser_tool._find_agent_browser", return_value="/usr/local/bin/agent-browser"), \
+             patch("tools.browser_tool._get_session_info", return_value=fake_session), \
+             patch("tools.browser_tool._socket_safe_tmpdir", return_value=str(tmp_path)), \
+             patch("tools.browser_tool._discover_homebrew_node_dirs", return_value=[]), \
+             patch("os.path.isdir", side_effect=selective_isdir), \
+             patch("subprocess.Popen", side_effect=capture_popen), \
+             patch("os.open", return_value=99), \
+             patch("os.close"), \
+             patch("tools.interrupt.is_interrupted", return_value=False), \
+             patch.dict(os.environ, {"PATH": "/usr/bin:/bin", "HOME": "/home/test"}, clear=True):
+            with patch("builtins.open", mock_open(read_data=fake_json)):
+                _run_browser_command("test-task", "navigate", ["https://example.com"])
+
+        result_path = captured_env.get("PATH", "")
+        assert "/data/data/com.termux/files/usr/bin" in result_path
+        assert "/data/data/com.termux/files/usr/sbin" in result_path

tests/tools/test_mcp_dynamic_discovery.py

@@ -21,34 +21,19 @@ class TestRegisterServerTools:
     def mock_registry(self):
         return ToolRegistry()
 
-    @pytest.fixture
-    def mock_toolsets(self):
-        return {
-            "hermes-cli": {"tools": ["terminal"], "description": "CLI", "includes": []},
-            "hermes-telegram": {"tools": ["terminal"], "description": "TG", "includes": []},
-            "custom-toolset": {"tools": [], "description": "Other", "includes": []},
-        }
-
-    def test_injects_hermes_toolsets(self, mock_registry, mock_toolsets):
-        """Tools are injected into hermes-* toolsets but not custom ones."""
+    def test_exposes_live_server_aliases(self, mock_registry):
+        """Registered MCP tools are reachable via live raw-server aliases."""
         server = MCPServerTask("my_srv")
         server._tools = [_make_mcp_tool("my_tool", "desc")]
         server.session = MagicMock()
+        from toolsets import resolve_toolset, validate_toolset
 
-        with patch("tools.registry.registry", mock_registry), \
-            patch("toolsets.create_custom_toolset"), \
-            patch.dict("toolsets.TOOLSETS", mock_toolsets, clear=True):
-
+        with patch("tools.registry.registry", mock_registry):
             registered = _register_server_tools("my_srv", server, {})
-
-        assert "mcp_my_srv_my_tool" in registered
-        assert "mcp_my_srv_my_tool" in mock_registry.get_all_tool_names()
-
-        # Injected into hermes-* toolsets
-        assert "mcp_my_srv_my_tool" in mock_toolsets["hermes-cli"]["tools"]
-        assert "mcp_my_srv_my_tool" in mock_toolsets["hermes-telegram"]["tools"]
-        # NOT into non-hermes toolsets
-        assert "mcp_my_srv_my_tool" not in mock_toolsets["custom-toolset"]["tools"]
+            assert "mcp_my_srv_my_tool" in registered
+            assert "mcp_my_srv_my_tool" in mock_registry.get_all_tool_names()
+            assert validate_toolset("my_srv") is True
+            assert "mcp_my_srv_my_tool" in resolve_toolset("my_srv")
 
 
 class TestRefreshTools:
@@ -58,19 +43,13 @@ class TestRefreshTools:
     def mock_registry(self):
         return ToolRegistry()
 
-    @pytest.fixture
-    def mock_toolsets(self):
-        return {
-            "hermes-cli": {"tools": ["terminal"], "description": "CLI", "includes": []},
-            "hermes-telegram": {"tools": ["terminal"], "description": "TG", "includes": []},
-        }
-
     @pytest.mark.asyncio
-    async def test_nuke_and_repave(self, mock_registry, mock_toolsets):
+    async def test_nuke_and_repave(self, mock_registry):
         """Old tools are removed and new tools registered on refresh."""
         server = MCPServerTask("live_srv")
         server._refresh_lock = asyncio.Lock()
         server._config = {}
+        from toolsets import resolve_toolset
 
         # Seed initial state: one old tool registered
         mock_registry.register(
@@ -79,7 +58,6 @@ class TestRefreshTools:
             description="", emoji="",
         )
         server._registered_tool_names = ["mcp_live_srv_old_tool"]
-        mock_toolsets["hermes-cli"]["tools"].append("mcp_live_srv_old_tool")
 
         # New tool list from server
         new_tool = _make_mcp_tool("new_tool", "new behavior")
@@ -89,20 +67,13 @@ class TestRefreshTools:
             )
         )
 
-        with patch("tools.registry.registry", mock_registry), \
-            patch("toolsets.create_custom_toolset"), \
-            patch.dict("toolsets.TOOLSETS", mock_toolsets, clear=True):
-
+        with patch("tools.registry.registry", mock_registry):
             await server._refresh_tools()
-
-        # Old tool completely gone
-        assert "mcp_live_srv_old_tool" not in mock_registry.get_all_tool_names()
-        assert "mcp_live_srv_old_tool" not in mock_toolsets["hermes-cli"]["tools"]
-
-        # New tool registered
-        assert "mcp_live_srv_new_tool" in mock_registry.get_all_tool_names()
-        assert "mcp_live_srv_new_tool" in mock_toolsets["hermes-cli"]["tools"]
-        assert server._registered_tool_names == ["mcp_live_srv_new_tool"]
+            assert "mcp_live_srv_old_tool" not in mock_registry.get_all_tool_names()
+            assert "mcp_live_srv_old_tool" not in resolve_toolset("live_srv")
+            assert "mcp_live_srv_new_tool" in mock_registry.get_all_tool_names()
+            assert "mcp_live_srv_new_tool" in resolve_toolset("live_srv")
+            assert server._registered_tool_names == ["mcp_live_srv_new_tool"]
 
 
 class TestMessageHandler:
@@ -165,6 +136,25 @@ class TestDeregister:
         # bar still in ts1, so check should remain
         assert "ts1" in reg._toolset_checks
 
+    def test_removes_toolset_alias_when_last_tool_is_removed(self):
+        reg = ToolRegistry()
+        reg.register(name="foo", toolset="mcp-srv", schema={}, handler=lambda x: x)
+        reg.register_toolset_alias("srv", "mcp-srv")
+
+        reg.deregister("foo")
+
+        assert reg.get_toolset_alias_target("srv") is None
+
+    def test_preserves_toolset_alias_while_toolset_still_exists(self):
+        reg = ToolRegistry()
+        reg.register(name="foo", toolset="mcp-srv", schema={}, handler=lambda x: x)
+        reg.register(name="bar", toolset="mcp-srv", schema={}, handler=lambda x: x)
+        reg.register_toolset_alias("srv", "mcp-srv")
+
+        reg.deregister("foo")
+
+        assert reg.get_toolset_alias_target("srv") == "mcp-srv"
+
     def test_noop_for_unknown_tool(self):
         reg = ToolRegistry()
         reg.deregister("nonexistent")  # Should not raise

tests/tools/test_mcp_tool.py

@@ -184,11 +184,7 @@ class TestToolHandler:
     def _patch_mcp_loop(self, coro_side_effect=None):
         """Return a patch for _run_on_mcp_loop that runs the coroutine directly."""
         def fake_run(coro, timeout=30):
-            loop = asyncio.new_event_loop()
-            try:
-                return loop.run_until_complete(coro)
-            finally:
-                loop.close()
+            return asyncio.run(coro)
         if coro_side_effect:
             return patch("tools.mcp_tool._run_on_mcp_loop", side_effect=coro_side_effect)
         return patch("tools.mcp_tool._run_on_mcp_loop", side_effect=fake_run)
@@ -365,10 +361,13 @@ class TestDiscoverAndRegister:
 
         _servers.pop("fs", None)
 
-    def test_toolset_created(self):
-        """A custom toolset is created for the MCP server."""
+    def test_toolset_resolves_live_from_registry(self):
+        """MCP toolsets resolve through the live registry without TOOLSETS mutation."""
+        from tools.registry import ToolRegistry
         from tools.mcp_tool import _discover_and_register_server, _servers, MCPServerTask
+        from toolsets import resolve_toolset, validate_toolset
 
+        mock_registry = ToolRegistry()
         mock_tools = [_make_mcp_tool("ping", "Ping")]
         mock_session = MagicMock()
 
@@ -378,16 +377,16 @@ class TestDiscoverAndRegister:
             server._tools = mock_tools
             return server
 
-        mock_create = MagicMock()
         with patch("tools.mcp_tool._connect_server", side_effect=fake_connect), \
-             patch("toolsets.create_custom_toolset", mock_create):
+             patch("tools.registry.registry", mock_registry):
             asyncio.run(
                 _discover_and_register_server("myserver", {"command": "test"})
             )
 
-        mock_create.assert_called_once()
-        call_kwargs = mock_create.call_args
-        assert call_kwargs[1]["name"] == "mcp-myserver" or call_kwargs[0][0] == "mcp-myserver"
+            assert validate_toolset("myserver") is True
+            assert validate_toolset("mcp-myserver") is True
+            assert "mcp_myserver_ping" in resolve_toolset("myserver")
+            assert "mcp_myserver_ping" in resolve_toolset("mcp-myserver")
 
         _servers.pop("myserver", None)
 
@@ -550,12 +549,15 @@ class TestMCPServerTask:
 # ---------------------------------------------------------------------------
 
 class TestToolsetInjection:
-    def test_mcp_tools_added_to_all_hermes_toolsets(self):
-        """Discovered MCP tools are dynamically injected into all hermes-* toolsets."""
+    def test_mcp_tools_resolve_through_server_aliases(self):
+        """Discovered MCP tools resolve through raw server-name aliases."""
         from tools.mcp_tool import MCPServerTask
+        from tools.registry import ToolRegistry
+        from toolsets import resolve_toolset, validate_toolset
 
         mock_tools = [_make_mcp_tool("list_files", "List files")]
         mock_session = MagicMock()
+        mock_registry = ToolRegistry()
 
         fresh_servers = {}
 
@@ -565,43 +567,32 @@ class TestToolsetInjection:
             server._tools = mock_tools
             return server
 
-        fake_toolsets = {
-            "hermes-cli": {"tools": ["terminal"], "description": "CLI", "includes": []},
-            "hermes-telegram": {"tools": ["terminal"], "description": "TG", "includes": []},
-            "hermes-gateway": {"tools": [], "description": "GW", "includes": []},
-            "non-hermes": {"tools": [], "description": "other", "includes": []},
-        }
         fake_config = {"fs": {"command": "npx", "args": []}}
 
         with patch("tools.mcp_tool._MCP_AVAILABLE", True), \
              patch("tools.mcp_tool._servers", fresh_servers), \
              patch("tools.mcp_tool._load_mcp_config", return_value=fake_config), \
              patch("tools.mcp_tool._connect_server", side_effect=fake_connect), \
-             patch("toolsets.TOOLSETS", fake_toolsets):
+             patch("tools.registry.registry", mock_registry):
             from tools.mcp_tool import discover_mcp_tools
             result = discover_mcp_tools()
 
-        assert "mcp_fs_list_files" in result
-        # All hermes-* toolsets get injection
-        assert "mcp_fs_list_files" in fake_toolsets["hermes-cli"]["tools"]
-        assert "mcp_fs_list_files" in fake_toolsets["hermes-telegram"]["tools"]
-        assert "mcp_fs_list_files" in fake_toolsets["hermes-gateway"]["tools"]
-        # Non-hermes toolset should NOT get injection
-        assert "mcp_fs_list_files" not in fake_toolsets["non-hermes"]["tools"]
-        # Original tools preserved
-        assert "terminal" in fake_toolsets["hermes-cli"]["tools"]
-        # Server name becomes a standalone toolset
-        assert "fs" in fake_toolsets
-        assert "mcp_fs_list_files" in fake_toolsets["fs"]["tools"]
-        assert fake_toolsets["fs"]["description"].startswith("MCP server '")
+            assert "mcp_fs_list_files" in result
+            assert validate_toolset("fs") is True
+            assert validate_toolset("mcp-fs") is True
+            assert "mcp_fs_list_files" in resolve_toolset("fs")
+            assert "mcp_fs_list_files" in resolve_toolset("mcp-fs")
 
     def test_server_toolset_skips_builtin_collision(self):
-        """MCP server named after a built-in toolset shouldn't overwrite it."""
+        """MCP raw aliases never overwrite a built-in toolset name."""
         from tools.mcp_tool import MCPServerTask
+        from tools.registry import ToolRegistry
+        from toolsets import resolve_toolset, validate_toolset
 
         mock_tools = [_make_mcp_tool("run", "Run command")]
         mock_session = MagicMock()
         fresh_servers = {}
+        mock_registry = ToolRegistry()
 
         async def fake_connect(name, config):
             server = MCPServerTask(name)
@@ -620,12 +611,15 @@ class TestToolsetInjection:
              patch("tools.mcp_tool._servers", fresh_servers), \
              patch("tools.mcp_tool._load_mcp_config", return_value=fake_config), \
              patch("tools.mcp_tool._connect_server", side_effect=fake_connect), \
+             patch("tools.registry.registry", mock_registry), \
              patch("toolsets.TOOLSETS", fake_toolsets):
             from tools.mcp_tool import discover_mcp_tools
             discover_mcp_tools()
 
-        # Built-in toolset preserved — description unchanged
-        assert fake_toolsets["terminal"]["description"] == "Terminal tools"
+            assert fake_toolsets["terminal"]["description"] == "Terminal tools"
+            assert "mcp_terminal_run" not in resolve_toolset("terminal")
+            assert validate_toolset("mcp-terminal") is True
+            assert "mcp_terminal_run" in resolve_toolset("mcp-terminal")
 
     def test_server_connection_failure_skipped(self):
         """If one server fails to connect, others still proceed."""
@@ -776,6 +770,42 @@ class TestShutdown:
         assert len(_servers) == 0
         mock_server.shutdown.assert_called_once()
 
+    def test_shutdown_deregisters_registered_tools(self):
+        """shutdown_mcp_servers removes MCP tools and their raw alias."""
+        import tools.mcp_tool as mcp_mod
+        from tools.mcp_tool import MCPServerTask, shutdown_mcp_servers, _servers
+        from tools.registry import registry
+        from toolsets import resolve_toolset, validate_toolset
+
+        _servers.clear()
+        registry.register(
+            name="mcp_test_ping",
+            toolset="mcp-test",
+            schema={
+                "name": "mcp_test_ping",
+                "description": "Ping",
+                "parameters": {"type": "object", "properties": {}},
+            },
+            handler=lambda *_args, **_kwargs: "{}",
+        )
+        registry.register_toolset_alias("test", "mcp-test")
+
+        server = MCPServerTask("test")
+        server._registered_tool_names = ["mcp_test_ping"]
+        _servers["test"] = server
+
+        mcp_mod._ensure_mcp_loop()
+        try:
+            assert validate_toolset("test") is True
+            assert "mcp_test_ping" in resolve_toolset("test")
+            shutdown_mcp_servers()
+        finally:
+            mcp_mod._mcp_loop = None
+            mcp_mod._mcp_thread = None
+
+        assert "mcp_test_ping" not in registry.get_all_tool_names()
+        assert validate_toolset("test") is False
+
     def test_shutdown_handles_errors(self):
         """shutdown_mcp_servers handles errors during close gracefully."""
         import tools.mcp_tool as mcp_mod
@@ -1179,7 +1209,11 @@ class TestConfigurableTimeouts:
         try:
             handler = _make_tool_handler("test_srv", "my_tool", 180)
             with patch("tools.mcp_tool._run_on_mcp_loop") as mock_run:
-                mock_run.return_value = json.dumps({"result": "ok"})
+                def fake_run(coro, timeout=30):
+                    coro.close()
+                    return json.dumps({"result": "ok"})
+
+                mock_run.side_effect = fake_run
                 handler({})
                 # Verify timeout=180 was passed
                 call_kwargs = mock_run.call_args
@@ -1279,11 +1313,7 @@ class TestUtilityHandlers:
     def _patch_mcp_loop(self):
         """Return a patch for _run_on_mcp_loop that runs the coroutine directly."""
         def fake_run(coro, timeout=30):
-            loop = asyncio.new_event_loop()
-            try:
-                return loop.run_until_complete(coro)
-            finally:
-                loop.close()
+            return asyncio.run(coro)
         return patch("tools.mcp_tool._run_on_mcp_loop", side_effect=fake_run)
 
     # -- list_resources --
@@ -3038,14 +3068,23 @@ class TestSanitizeMcpNameComponent:
             assert "/" not in name
             assert "." not in name
 
-    def test_slash_in_sync_mcp_toolsets(self):
-        """_sync_mcp_toolsets uses sanitize consistently with _convert_mcp_schema."""
-        from tools.mcp_tool import sanitize_mcp_name_component
+    def test_slash_in_server_alias_resolution(self):
+        """Server names with slashes resolve through their live MCP alias."""
+        from tools.registry import ToolRegistry
+        from toolsets import resolve_toolset, validate_toolset
 
-        # Verify the prefix generation matches what _convert_mcp_schema produces
-        server_name = "ai.exa/exa"
-        safe_prefix = f"mcp_{sanitize_mcp_name_component(server_name)}_"
-        assert safe_prefix == "mcp_ai_exa_exa_"
+        reg = ToolRegistry()
+        reg.register(
+            name="mcp_ai_exa_exa_search",
+            toolset="mcp-ai.exa/exa",
+            schema={"name": "mcp_ai_exa_exa_search", "description": "Search", "parameters": {"type": "object", "properties": {}}},
+            handler=lambda *_args, **_kwargs: "{}",
+        )
+        reg.register_toolset_alias("ai.exa/exa", "mcp-ai.exa/exa")
+
+        with patch("tools.registry.registry", reg):
+            assert validate_toolset("ai.exa/exa") is True
+            assert "mcp_ai_exa_exa_search" in resolve_toolset("ai.exa/exa")
 
 
 # ---------------------------------------------------------------------------