diff --git a/ROADMAP.md b/ROADMAP.md
index 41cfb98b..02d6af2c 100644
--- a/ROADMAP.md
+++ b/ROADMAP.md
@@ -168,7 +168,7 @@ Remaining gaps and forward work live in [Forward Work](#forward-work) below.
 
 ### Slash commands
 - [x] Command registry + autocomplete dropdown
-- [x] Built-ins: `/help`, `/clear`, `/model`, `/workspace`, `/new`, `/usage`, `/theme`, `/compact`, `/queue`, `/interrupt`, `/steer`, `/btw`, `/reasoning`, `/skills`, `/toolsets`
+- [x] Built-ins: `/help`, `/clear`, `/model`, `/workspace`, `/new`, `/usage`, `/theme`, `/compact`, `/queue`, `/interrupt`, `/steer`, `/goal`, `/btw`, `/reasoning`, `/skills`, `/toolsets`
 - [x] Transparent pass-through for unrecognized commands
 
 ### Security
diff --git a/api/goals.py b/api/goals.py
new file mode 100644
index 00000000..32d0262d
--- /dev/null
+++ b/api/goals.py
@@ -0,0 +1,489 @@
+"""WebUI bridge for Hermes persistent session goals."""
+
+from __future__ import annotations
+
+import copy
+import logging
+import time
+from pathlib import Path
+from typing import Any, Dict, Optional
+
+logger = logging.getLogger(__name__)
+
+try:  # Exposed as a module attribute so tests can monkeypatch it directly.
+    from hermes_cli.goals import (  # type: ignore
+        CONTINUATION_PROMPT_TEMPLATE,
+        DEFAULT_MAX_TURNS,
+        GoalManager as _NativeGoalManager,
+        GoalState,
+        judge_goal,
+    )
+except Exception:  # pragma: no cover - depends on installed hermes-agent
+    CONTINUATION_PROMPT_TEMPLATE = ""  # type: ignore
+    DEFAULT_MAX_TURNS = 20  # type: ignore
+    _NativeGoalManager = None  # type: ignore
+    GoalState = None  # type: ignore
+    judge_goal = None  # type: ignore
+
+GoalManager = _NativeGoalManager  # type: ignore
+
+_DB_CACHE: dict[str, Any] = {}
+
+
+def _default_max_turns() -> int:
+    """Return the configured /goal turn budget, defaulting to Hermes' 20 turns."""
+    try:
+        from api import config as _config
+
+        cfg = getattr(_config, "cfg", {}) or {}
+        goals_cfg = cfg.get("goals", {}) if isinstance(cfg, dict) else {}
+        if not isinstance(goals_cfg, dict):
+            return int(DEFAULT_MAX_TURNS or 20)
+        return max(1, int(goals_cfg.get("max_turns", DEFAULT_MAX_TURNS or 20) or 20))
+    except Exception:
+        return int(DEFAULT_MAX_TURNS or 20)
+
+
+def _meta_key(session_id: str) -> str:
+    return f"goal:{session_id}"
+
+
+def _profile_db(profile_home: str | Path):
+    """Return a SessionDB pinned to *profile_home*, without reading HERMES_HOME.
+
+    The upstream Hermes GoalManager persists through hermes_cli.goals.load_goal(),
+    which resolves SessionDB from process-global HERMES_HOME. WebUI sessions are
+    profile-scoped and can run concurrently, so the WebUI bridge uses an explicit
+    state.db path whenever the caller provides the session's profile home.
+    """
+    home = Path(profile_home).expanduser().resolve()
+    key = str(home)
+    cached = _DB_CACHE.get(key)
+    if cached is not None:
+        return cached
+    try:
+        from hermes_state import SessionDB  # type: ignore
+
+        db = SessionDB(db_path=home / "state.db")
+    except Exception as exc:  # pragma: no cover - import/env dependent
+        logger.debug("GoalManager profile DB unavailable for %s: %s", home, exc)
+        return None
+    _DB_CACHE[key] = db
+    return db
+
+
+class _ProfileGoalManager:
+    """Small WebUI-local GoalManager adapter with explicit profile persistence."""
+
+    def __init__(self, session_id: str, *, profile_home: str | Path, default_max_turns: int = 20):
+        if GoalState is None:
+            raise RuntimeError("Hermes goal state unavailable")
+        self.session_id = session_id
+        self.profile_home = Path(profile_home).expanduser().resolve()
+        self.default_max_turns = int(default_max_turns or DEFAULT_MAX_TURNS or 20)
+        self._state = self._load()
+
+    @property
+    def state(self):
+        return self._state
+
+    def _load(self):
+        db = _profile_db(self.profile_home)
+        if db is None or not self.session_id:
+            return None
+        try:
+            raw = db.get_meta(_meta_key(self.session_id))
+        except Exception as exc:
+            logger.debug("GoalManager profile get_meta failed: %s", exc)
+            return None
+        if not raw:
+            return None
+        try:
+            return GoalState.from_json(raw)  # type: ignore[union-attr]
+        except Exception as exc:
+            logger.warning("GoalManager profile state parse failed for %s: %s", self.session_id, exc)
+            return None
+
+    def _save(self, state) -> None:
+        db = _profile_db(self.profile_home)
+        if db is None or not self.session_id or state is None:
+            return
+        try:
+            db.set_meta(_meta_key(self.session_id), state.to_json())
+        except Exception as exc:
+            logger.debug("GoalManager profile set_meta failed: %s", exc)
+
+    def is_active(self) -> bool:
+        return self._state is not None and self._state.status == "active"
+
+    def has_goal(self) -> bool:
+        return self._state is not None and self._state.status in ("active", "paused")
+
+    def status_line(self) -> str:
+        s = self._state
+        if s is None or s.status in ("cleared",):
+            return "No active goal. Set one with /goal <text>."
+        turns = f"{s.turns_used}/{s.max_turns} turns"
+        if s.status == "active":
+            return f"⊙ Goal (active, {turns}): {s.goal}"
+        if s.status == "paused":
+            extra = f" — {s.paused_reason}" if s.paused_reason else ""
+            return f"⏸ Goal (paused, {turns}{extra}): {s.goal}"
+        if s.status == "done":
+            return f"✓ Goal done ({turns}): {s.goal}"
+        return f"Goal ({s.status}, {turns}): {s.goal}"
+
+    def set(self, goal: str, *, max_turns: Optional[int] = None):
+        goal = (goal or "").strip()
+        if not goal:
+            raise ValueError("goal text is empty")
+        state = GoalState(  # type: ignore[operator]
+            goal=goal,
+            status="active",
+            turns_used=0,
+            max_turns=int(max_turns) if max_turns else self.default_max_turns,
+            created_at=time.time(),
+            last_turn_at=0.0,
+        )
+        self._state = state
+        self._save(state)
+        return state
+
+    def pause(self, reason: str = "user-paused"):
+        if not self._state:
+            return None
+        self._state.status = "paused"
+        self._state.paused_reason = reason
+        self._save(self._state)
+        return self._state
+
+    def resume(self, *, reset_budget: bool = True):
+        if not self._state:
+            return None
+        self._state.status = "active"
+        self._state.paused_reason = None
+        if reset_budget:
+            self._state.turns_used = 0
+        self._save(self._state)
+        return self._state
+
+    def clear(self) -> None:
+        if self._state is None:
+            return
+        self._state.status = "cleared"
+        self._save(self._state)
+        self._state = None
+
+    def evaluate_after_turn(self, last_response: str, *, user_initiated: bool = True) -> Dict[str, Any]:
+        state = self._state
+        if state is None or state.status != "active":
+            return {
+                "status": state.status if state else None,
+                "should_continue": False,
+                "continuation_prompt": None,
+                "verdict": "inactive",
+                "reason": "no active goal",
+                "message": "",
+            }
+
+        state.turns_used += 1
+        state.last_turn_at = time.time()
+
+        if judge_goal is None:
+            verdict, reason = "continue", "goal judge unavailable"
+        else:
+            verdict, reason = judge_goal(state.goal, str(last_response or ""))
+        state.last_verdict = verdict
+        state.last_reason = reason
+
+        if verdict == "done":
+            state.status = "done"
+            self._save(state)
+            return {
+                "status": "done",
+                "should_continue": False,
+                "continuation_prompt": None,
+                "verdict": "done",
+                "reason": reason,
+                "message": f"✓ Goal achieved: {reason}",
+            }
+
+        if state.turns_used >= state.max_turns:
+            state.status = "paused"
+            state.paused_reason = f"turn budget exhausted ({state.turns_used}/{state.max_turns})"
+            self._save(state)
+            return {
+                "status": "paused",
+                "should_continue": False,
+                "continuation_prompt": None,
+                "verdict": "continue",
+                "reason": reason,
+                "message": (
+                    f"⏸ Goal paused — {state.turns_used}/{state.max_turns} turns used. "
+                    "Use /goal resume to keep going, or /goal clear to stop."
+                ),
+            }
+
+        self._save(state)
+        return {
+            "status": "active",
+            "should_continue": True,
+            "continuation_prompt": self.next_continuation_prompt(),
+            "verdict": "continue",
+            "reason": reason,
+            "message": f"↻ Continuing toward goal ({state.turns_used}/{state.max_turns}): {reason}",
+        }
+
+    def next_continuation_prompt(self) -> Optional[str]:
+        if not self._state or self._state.status != "active":
+            return None
+        return CONTINUATION_PROMPT_TEMPLATE.format(goal=self._state.goal)
+
+
+def _manager(session_id: str, *, profile_home: str | Path | None = None):
+    if GoalManager is None:
+        return None
+    if profile_home and GoalManager is _NativeGoalManager and GoalState is not None:
+        try:
+            return _ProfileGoalManager(
+                session_id=session_id,
+                profile_home=profile_home,
+                default_max_turns=_default_max_turns(),
+            )
+        except Exception as exc:
+            logger.debug("Profile-scoped GoalManager unavailable: %s", exc)
+            return None
+    return GoalManager(session_id=session_id, default_max_turns=_default_max_turns())
+
+
+def _state_payload(state: Any) -> Optional[Dict[str, Any]]:
+    if state is None:
+        return None
+    return {
+        "goal": getattr(state, "goal", "") or "",
+        "status": getattr(state, "status", "") or "",
+        "turns_used": int(getattr(state, "turns_used", 0) or 0),
+        "max_turns": int(getattr(state, "max_turns", 0) or 0),
+        "last_verdict": getattr(state, "last_verdict", None),
+        "last_reason": getattr(state, "last_reason", None),
+        "paused_reason": getattr(state, "paused_reason", None),
+    }
+
+
+def _payload(
+    *,
+    ok: bool = True,
+    action: str,
+    message: str,
+    state: Any = None,
+    error: str | None = None,
+    kickoff_prompt: str | None = None,
+    decision: Dict[str, Any] | None = None,
+) -> Dict[str, Any]:
+    body: Dict[str, Any] = {
+        "ok": bool(ok),
+        "action": action,
+        "message": message,
+        "goal": _state_payload(state),
+    }
+    if error:
+        body["error"] = error
+    if kickoff_prompt:
+        body["kickoff_prompt"] = kickoff_prompt
+    if decision is not None:
+        body["decision"] = decision
+    return body
+
+
+def goal_state_snapshot(session_id: str, *, profile_home: str | Path | None = None) -> Any:
+    """Return a deep copy of current goal state for rollback before kickoff."""
+    mgr = _manager(str(session_id or ""), profile_home=profile_home)
+    if mgr is None:
+        return None
+    return copy.deepcopy(getattr(mgr, "state", None))
+
+
+def restore_goal_state(session_id: str, snapshot: Any, *, profile_home: str | Path | None = None) -> None:
+    """Restore a prior goal state after kickoff stream creation fails."""
+    mgr = _manager(str(session_id or ""), profile_home=profile_home)
+    if mgr is None:
+        return
+    if snapshot is None:
+        try:
+            mgr.clear()
+        except Exception:
+            pass
+        return
+    if isinstance(mgr, _ProfileGoalManager):
+        mgr._state = snapshot
+        mgr._save(snapshot)
+        return
+    try:
+        from hermes_cli.goals import save_goal  # type: ignore
+
+        save_goal(str(session_id or ""), snapshot)
+    except Exception as exc:  # pragma: no cover - native fallback only
+        logger.debug("Goal state restore failed for %s: %s", session_id, exc)
+
+
+def goal_command_payload(
+    session_id: str,
+    args: str = "",
+    *,
+    stream_running: bool = False,
+    profile_home: str | Path | None = None,
+) -> Dict[str, Any]:
+    """Return the WebUI response payload for a /goal command.
+
+    Mirrors the gateway command semantics:
+    - /goal or /goal status shows status
+    - /goal pause pauses
+    - /goal resume resumes without auto-starting a turn
+    - /goal clear|stop|done clears
+    - /goal <text> sets a new active goal and returns kickoff_prompt so the
+      caller can start the first normal user-role turn immediately.
+    """
+    sid = str(session_id or "").strip()
+    if not sid:
+        return _payload(ok=False, action="error", error="missing_session", message="session_id required")
+
+    mgr = _manager(sid, profile_home=profile_home)
+    if mgr is None:
+        return _payload(ok=False, action="error", error="unavailable", message="Goals unavailable on this session.")
+
+    text = str(args or "").strip()
+    lower = text.lower()
+
+    if not text or lower == "status":
+        return _payload(action="status", message=mgr.status_line(), state=getattr(mgr, "state", None))
+
+    if lower == "pause":
+        state = mgr.pause(reason="user-paused")
+        if state is None:
+            return _payload(ok=False, action="pause", error="no_goal", message="No goal set.")
+        return _payload(action="pause", message=f"⏸ Goal paused: {state.goal}", state=state)
+
+    if lower == "resume":
+        state = mgr.resume()
+        if state is None:
+            return _payload(ok=False, action="resume", error="no_goal", message="No goal to resume.")
+        return _payload(
+            action="resume",
+            message=(
+                f"▶ Goal resumed: {state.goal}\n"
+                "Send a new message, or type continue, to kick it off."
+            ),
+            state=state,
+        )
+
+    if lower in ("clear", "stop", "done"):
+        had = bool(mgr.has_goal())
+        mgr.clear()
+        return _payload(
+            action="clear",
+            message="Goal cleared." if had else "No active goal.",
+            state=getattr(mgr, "state", None),
+        )
+
+    if stream_running:
+        return _payload(
+            ok=False,
+            action="set",
+            error="agent_running",
+            message=(
+                "Agent is running — use /goal status / pause / clear mid-run, "
+                "or /stop before setting a new goal."
+            ),
+        )
+
+    try:
+        state = mgr.set(text)
+    except ValueError as exc:
+        return _payload(ok=False, action="set", error="invalid_goal", message=f"Invalid goal: {exc}")
+
+    return _payload(
+        action="set",
+        message=(
+            f"⊙ Goal set ({state.max_turns}-turn budget): {state.goal}\n"
+            "I'll keep working until the goal is done, you pause/clear it, or the budget is exhausted.\n"
+            "Controls: /goal status · /goal pause · /goal resume · /goal clear"
+        ),
+        state=state,
+        kickoff_prompt=state.goal,
+    )
+
+
+def has_active_goal(
+    session_id: str,
+    *,
+    profile_home: str | Path | None = None,
+) -> bool:
+    """Return True when the session has an active standing goal to evaluate."""
+    sid = str(session_id or "").strip()
+    if not sid:
+        return False
+    mgr = _manager(sid, profile_home=profile_home)
+    if mgr is None:
+        return False
+    try:
+        return bool(mgr.is_active())
+    except Exception as exc:
+        logger.debug("goal active-state check failed for session=%s: %s", sid, exc)
+        return False
+
+
+def evaluate_goal_after_turn(
+    session_id: str,
+    last_response: str,
+    *,
+    user_initiated: bool = True,
+    profile_home: str | Path | None = None,
+) -> Dict[str, Any]:
+    """Evaluate a completed turn against the standing goal, if any."""
+    sid = str(session_id or "").strip()
+    if not sid:
+        return {
+            "status": None,
+            "should_continue": False,
+            "continuation_prompt": None,
+            "verdict": "inactive",
+            "reason": "missing session_id",
+            "message": "",
+        }
+    mgr = _manager(sid, profile_home=profile_home)
+    if mgr is None:
+        return {
+            "status": None,
+            "should_continue": False,
+            "continuation_prompt": None,
+            "verdict": "inactive",
+            "reason": "goals unavailable",
+            "message": "",
+        }
+    try:
+        if not mgr.is_active():
+            return {
+                "status": getattr(getattr(mgr, "state", None), "status", None),
+                "should_continue": False,
+                "continuation_prompt": None,
+                "verdict": "inactive",
+                "reason": "no active goal",
+                "message": "",
+            }
+        decision = mgr.evaluate_after_turn(str(last_response or ""), user_initiated=user_initiated)
+    except Exception as exc:
+        logger.debug("goal evaluation failed for session=%s: %s", sid, exc)
+        return {
+            "status": None,
+            "should_continue": False,
+            "continuation_prompt": None,
+            "verdict": "error",
+            "reason": f"goal evaluation failed: {type(exc).__name__}",
+            "message": "",
+        }
+    if not isinstance(decision, dict):
+        decision = {}
+    decision.setdefault("should_continue", False)
+    decision.setdefault("continuation_prompt", None)
+    decision.setdefault("message", "")
+    return decision
diff --git a/api/routes.py b/api/routes.py
index 151b03a8..26bcd782 100644
--- a/api/routes.py
+++ b/api/routes.py
@@ -4269,6 +4269,9 @@ def handle_post(handler, parsed) -> bool:
     if parsed.path == "/api/background":
         return _handle_background(handler, body)
 
+    if parsed.path == "/api/goal":
+        return _handle_goal_command(handler, body)
+
     if parsed.path == "/api/chat/start":
         return _handle_chat_start(handler, body, diag=diag)
 
@@ -6438,6 +6441,197 @@ def _prepare_chat_start_session_for_stream(
     s.save()
 
 
+def _start_chat_stream_for_session(
+    s,
+    *,
+    msg: str,
+    attachments=None,
+    workspace: str,
+    model: str,
+    model_provider=None,
+    normalized_model: bool = False,
+    diag=None,
+):
+    """Persist pending state, register an SSE channel, and start an agent turn."""
+    attachments = attachments or []
+    # Prevent duplicate runs in the same session while a stream is still active.
+    # This commonly happens after page refresh/reconnect races and can produce
+    # duplicated clarify cards for what appears to be a single user request.
+    diag.stage("active_stream_check") if diag else None
+    current_stream_id = getattr(s, "active_stream_id", None)
+    if current_stream_id:
+        diag.stage("active_stream_lock_wait") if diag else None
+        with STREAMS_LOCK:
+            current_active = current_stream_id in STREAMS
+        if current_active:
+            diag.stage("response_write") if diag else None
+            return {
+                "error": "session already has an active stream",
+                "active_stream_id": current_stream_id,
+                "_status": 409,
+            }
+        # Stale stream id from a previous run; clear and continue.
+        diag.stage("stale_stream_cleanup") if diag else None
+        _clear_stale_stream_state(s)
+    stream_id = uuid.uuid4().hex
+    session_lock = _get_session_agent_lock(s.session_id)
+    diag.stage("session_lock_wait") if diag else None
+    with session_lock:
+        diag.stage("save_pending_state") if diag else None
+        _prepare_chat_start_session_for_stream(
+            s,
+            msg=msg,
+            attachments=attachments,
+            workspace=workspace,
+            model=model,
+            model_provider=model_provider,
+            stream_id=stream_id,
+        )
+    diag.stage("set_last_workspace") if diag else None
+    set_last_workspace(workspace)
+    diag.stage("stream_registration") if diag else None
+    stream = create_stream_channel()
+    with STREAMS_LOCK:
+        STREAMS[stream_id] = stream
+    diag.stage("worker_thread_start") if diag else None
+    thr = threading.Thread(
+        target=_run_agent_streaming,
+        args=(s.session_id, msg, model, workspace, stream_id, attachments),
+        kwargs={"model_provider": model_provider},
+        daemon=True,
+    )
+    thr.start()
+    response = {
+        "stream_id": stream_id,
+        "session_id": s.session_id,
+        "pending_started_at": s.pending_started_at,
+    }
+    if normalized_model:
+        response["effective_model"] = model
+    if model_provider:
+        response["effective_model_provider"] = model_provider
+    return response
+
+
+def _handle_goal_command(handler, body):
+    """Handle WebUI /goal command controls and optional kickoff stream."""
+    try:
+        require(body, "session_id")
+    except ValueError as e:
+        return bad(handler, str(e))
+    try:
+        s = get_session(body["session_id"])
+    except KeyError:
+        return bad(handler, "Session not found", 404)
+
+    requested_profile = str(body.get("profile") or "").strip()
+    if requested_profile:
+        try:
+            from api.profiles import _PROFILE_ID_RE
+
+            if requested_profile != "default" and not _PROFILE_ID_RE.fullmatch(requested_profile):
+                return bad(handler, "invalid profile", 400)
+        except ImportError:
+            requested_profile = ""
+    if requested_profile and not _profiles_match(getattr(s, "profile", None), requested_profile):
+        has_persisted_turns = bool(
+            getattr(s, "messages", None)
+            or getattr(s, "context_messages", None)
+            or getattr(s, "pending_user_message", None)
+        )
+        if not has_persisted_turns:
+            s.profile = requested_profile
+
+    current_stream_id = getattr(s, "active_stream_id", None)
+    stream_running = False
+    if current_stream_id:
+        with STREAMS_LOCK:
+            stream_running = current_stream_id in STREAMS
+        if not stream_running:
+            _clear_stale_stream_state(s)
+
+    try:
+        from api.profiles import get_hermes_home_for_profile
+
+        profile_home = get_hermes_home_for_profile(getattr(s, "profile", None))
+    except Exception:
+        profile_home = None
+
+    from api.goals import goal_command_payload, goal_state_snapshot, restore_goal_state
+
+    goal_args = str(body.get("args", "") or body.get("text", "") or "")
+    goal_action = goal_args.strip().lower()
+    will_kickoff = bool(
+        goal_args.strip()
+        and goal_action not in ("status", "pause", "resume", "clear", "stop", "done")
+        and not stream_running
+    )
+    workspace = model = model_provider = normalized_model = None
+    previous_goal_state = None
+    if will_kickoff:
+        try:
+            workspace = str(resolve_trusted_workspace(body.get("workspace") or s.workspace))
+        except ValueError as e:
+            return bad(handler, str(e))
+        requested_model = body.get("model") or s.model
+        requested_provider = (
+            body.get("model_provider")
+            if "model_provider" in body
+            else getattr(s, "model_provider", None)
+        )
+        model, model_provider, normalized_model = _resolve_compatible_session_model_state(
+            requested_model,
+            requested_provider,
+        )
+        previous_goal_state = goal_state_snapshot(s.session_id, profile_home=profile_home)
+
+    payload = goal_command_payload(
+        s.session_id,
+        goal_args,
+        stream_running=stream_running,
+        profile_home=profile_home,
+    )
+    if not payload.get("ok", True):
+        status = 409 if payload.get("error") == "agent_running" else 400
+        return j(handler, payload, status=status)
+
+    kickoff_prompt = str(payload.get("kickoff_prompt") or "").strip()
+    if kickoff_prompt:
+        if workspace is None:
+            try:
+                workspace = str(resolve_trusted_workspace(body.get("workspace") or s.workspace))
+            except ValueError as e:
+                return bad(handler, str(e))
+        if model is None:
+            requested_model = body.get("model") or s.model
+            requested_provider = (
+                body.get("model_provider")
+                if "model_provider" in body
+                else getattr(s, "model_provider", None)
+            )
+            model, model_provider, normalized_model = _resolve_compatible_session_model_state(
+                requested_model,
+                requested_provider,
+            )
+        stream_response = _start_chat_stream_for_session(
+            s,
+            msg=kickoff_prompt,
+            attachments=[],
+            workspace=workspace,
+            model=model,
+            model_provider=model_provider,
+            normalized_model=normalized_model,
+        )
+        status = int(stream_response.pop("_status", 200) or 200)
+        payload.update(stream_response)
+        if status >= 400:
+            restore_goal_state(s.session_id, previous_goal_state, profile_home=profile_home)
+            payload["ok"] = False
+            return j(handler, payload, status=status)
+
+    return j(handler, payload)
+
+
 def _handle_chat_start(handler, body, diag=None):
     try:
         diag.stage("validate_session_id") if diag else None
@@ -6494,72 +6688,25 @@ def _handle_chat_start(handler, body, diag=None):
             requested_model,
             requested_provider,
         )
-        # Prevent duplicate runs in the same session while a stream is still active.
-        # This commonly happens after page refresh/reconnect races and can produce
-        # duplicated clarify cards for what appears to be a single user request.
-        diag.stage("active_stream_check") if diag else None
-        current_stream_id = getattr(s, "active_stream_id", None)
-        if current_stream_id:
-            diag.stage("active_stream_lock_wait") if diag else None
-            with STREAMS_LOCK:
-                current_active = current_stream_id in STREAMS
-            if current_active:
-                diag.stage("response_write") if diag else None
-                return j(
-                    handler,
-                    {
-                        "error": "session already has an active stream",
-                        "active_stream_id": current_stream_id,
-                    },
-                    status=409,
-                )
-            # Stale stream id from a previous run; clear and continue.
-            diag.stage("stale_stream_cleanup") if diag else None
-            _clear_stale_stream_state(s)
-        stream_id = uuid.uuid4().hex
-        session_lock = _get_session_agent_lock(s.session_id)
-        diag.stage("session_lock_wait") if diag else None
-        with session_lock:
-            diag.stage("save_pending_state") if diag else None
-            _prepare_chat_start_session_for_stream(
-                s,
-                msg=msg,
-                attachments=attachments,
-                workspace=workspace,
-                model=model,
-                model_provider=model_provider,
-                stream_id=stream_id,
-            )
-        diag.stage("set_last_workspace") if diag else None
-        set_last_workspace(workspace)
-        diag.stage("stream_registration") if diag else None
-        stream = create_stream_channel()
-        with STREAMS_LOCK:
-            STREAMS[stream_id] = stream
-        diag.stage("worker_thread_start") if diag else None
-        thr = threading.Thread(
-            target=_run_agent_streaming,
-            args=(s.session_id, msg, model, workspace, stream_id, attachments),
-            kwargs={"model_provider": model_provider},
-            daemon=True,
+        response = _start_chat_stream_for_session(
+            s,
+            msg=msg,
+            attachments=attachments,
+            workspace=workspace,
+            model=model,
+            model_provider=model_provider,
+            normalized_model=normalized_model,
+            diag=diag,
         )
-        thr.start()
-        response = {
-            "stream_id": stream_id,
-            "session_id": s.session_id,
-            "pending_started_at": s.pending_started_at,
-        }
-        if normalized_model:
-            response["effective_model"] = model
-        if model_provider:
-            response["effective_model_provider"] = model_provider
+        status = int(response.pop("_status", 200) or 200)
         diag.stage("response_write") if diag else None
-        return j(handler, response)
+        return j(handler, response, status=status)
     finally:
         if diag:
             diag.finish()
 
 
+
 def _normalize_chat_attachments(raw_attachments):
     """Normalize attachment payloads from the browser.
 
diff --git a/api/streaming.py b/api/streaming.py
index c71fe074..d9daeb33 100644
--- a/api/streaming.py
+++ b/api/streaming.py
@@ -3227,6 +3227,64 @@ def _run_agent_streaming(
                     })
             except Exception:
                 logger.debug("Failed to drain pending steer for session %s", session_id)
+            # /goal parity: after a successful assistant turn, run the Hermes
+            # GoalManager judge before terminal done/stream_end events. The
+            # frontend surfaces the status line and queues continuation_prompt as
+            # a normal next user message so /queue and user input keep priority.
+            try:
+                from api.goals import evaluate_goal_after_turn, has_active_goal
+
+                if not has_active_goal(session_id, profile_home=_profile_home):
+                    _goal_decision = {}
+                else:
+                    _last_goal_response = ''
+                    for _goal_msg in reversed(s.messages or []):
+                        if not isinstance(_goal_msg, dict) or _goal_msg.get('role') != 'assistant':
+                            continue
+                        _goal_content = _goal_msg.get('content', '')
+                        if isinstance(_goal_content, list):
+                            _goal_parts = []
+                            for _goal_part in _goal_content:
+                                if isinstance(_goal_part, dict):
+                                    _goal_text = _goal_part.get('text') or _goal_part.get('content')
+                                    if _goal_text:
+                                        _goal_parts.append(str(_goal_text))
+                            _last_goal_response = '\n'.join(_goal_parts)
+                        else:
+                            _last_goal_response = str(_goal_content or '')
+                        break
+                    put('goal', {
+                        'session_id': session_id,
+                        'state': 'evaluating',
+                        'message': 'Evaluating goal progress…',
+                    })
+                    _goal_decision = evaluate_goal_after_turn(
+                        session_id,
+                        _last_goal_response,
+                        user_initiated=True,
+                        profile_home=_profile_home,
+                    )
+                decision = _goal_decision or {}
+                _goal_message = str(decision.get('message') or '').strip()
+                if _goal_message:
+                    put('goal', {
+                        'session_id': session_id,
+                        'state': 'continuing' if decision.get('should_continue') else 'idle',
+                        'message': _goal_message,
+                        'decision': decision,
+                    })
+                if decision.get('should_continue'):
+                    continuation_prompt = str(decision.get('continuation_prompt') or '').strip()
+                    if continuation_prompt:
+                        put('goal_continue', {
+                            'session_id': session_id,
+                            'continuation_prompt': continuation_prompt,
+                            'text': continuation_prompt,
+                            'message': _goal_message,
+                            'decision': decision,
+                        })
+            except Exception as _goal_exc:
+                logger.debug("Goal continuation hook failed for session %s: %s", session_id, _goal_exc)
             raw_session = s.compact() | {'messages': s.messages, 'tool_calls': tool_calls}
             put('done', {'session': redact_session_data(raw_session), 'usage': usage})
             # Emit one last metering packet for the live message-header TPS label.
diff --git a/docs/pr-media/1808/goal-autocomplete.png b/docs/pr-media/1808/goal-autocomplete.png
new file mode 100644
index 00000000..e07e32a5
Binary files /dev/null and b/docs/pr-media/1808/goal-autocomplete.png differ
diff --git a/docs/pr-media/1808/goal-command-set.png b/docs/pr-media/1808/goal-command-set.png
new file mode 100644
index 00000000..f5ee966d
Binary files /dev/null and b/docs/pr-media/1808/goal-command-set.png differ
diff --git a/docs/pr-media/1808/goal-status-command.png b/docs/pr-media/1808/goal-status-command.png
new file mode 100644
index 00000000..26eef14b
Binary files /dev/null and b/docs/pr-media/1808/goal-status-command.png differ
diff --git a/docs/pr-media/1866/goal-evaluating-status.png b/docs/pr-media/1866/goal-evaluating-status.png
new file mode 100644
index 00000000..a411a43a
Binary files /dev/null and b/docs/pr-media/1866/goal-evaluating-status.png differ
diff --git a/static/commands.js b/static/commands.js
index 90fc5aca..3d5aa9f0 100644
--- a/static/commands.js
+++ b/static/commands.js
@@ -18,6 +18,7 @@ const COMMANDS=[
   {name:'personality', desc:t('cmd_personality'), fn:cmdPersonality, arg:'name', subArgs:'personalities'},
   {name:'skills',    desc:t('cmd_skills'),   fn:cmdSkills,   arg:'query'},
   {name:'stop',      desc:t('cmd_stop'),     fn:cmdStop,      noEcho:true},
+  {name:'goal',      desc:t('cmd_goal'),     fn:cmdGoal,      arg:'[status|pause|resume|clear|text]', subArgs:['status','pause','resume','clear']},
   {name:'queue',     desc:t('cmd_queue'),    fn:cmdQueue,     arg:'message', noEcho:true},
   {name:'interrupt', desc:t('cmd_interrupt'), fn:cmdInterrupt, arg:'message', noEcho:true},
   {name:'steer',     desc:t('cmd_steer'),    fn:cmdSteer,     arg:'message', noEcho:true},
@@ -625,6 +626,53 @@ async function cmdStop(){
   else showToast(t('cancel_unavailable'));
 }
 
+async function cmdGoal(args){
+  if(!S.session){await newSession();await renderSessionList();}
+  if(!S.session||!S.session.session_id){showToast(t('no_active_session'));return;}
+  const activeSid=S.session.session_id;
+  try{
+    const r=await api('/api/goal',{method:'POST',body:JSON.stringify({
+      session_id:activeSid,
+      args:args||'',
+      workspace:S.session.workspace,
+      model:S.session.model||($('modelSelect')&&$('modelSelect').value)||'',
+      model_provider:S.session.model_provider||null,
+      profile:S.activeProfile||S.session.profile||'default',
+    })});
+    const msg=String((r&&r.message)||'').trim();
+    if(msg){
+      S.messages.push({role:'assistant',content:msg,_ts:Date.now()/1000,_goalStatus:true,_transient:true});
+      renderMessages({preserveScroll:true});
+      showToast(msg.split('\n')[0],2600);
+    }
+    if(!r||!r.stream_id)return;
+    S.toolCalls=[];
+    if(typeof clearLiveToolCards==='function')clearLiveToolCards();
+    appendThinking();setBusy(true);
+    setComposerStatus('Working toward goal…');
+    S.activeStreamId=r.stream_id;
+    if(S.session&&S.session.session_id===activeSid){
+      S.session.active_stream_id=r.stream_id;
+      if(typeof r.pending_started_at==='number')S.session.pending_started_at=r.pending_started_at;
+      if(r.effective_model)S.session.model=r.effective_model;
+      if(r.effective_model_provider)S.session.model_provider=r.effective_model_provider;
+    }
+    INFLIGHT[activeSid]={messages:[...S.messages],uploaded:[],toolCalls:[]};
+    if(typeof markInflight==='function')markInflight(activeSid,r.stream_id);
+    if(typeof saveInflightState==='function')saveInflightState(activeSid,{streamId:r.stream_id,messages:INFLIGHT[activeSid].messages,uploaded:[],toolCalls:[]});
+    startApprovalPolling(activeSid);
+    startClarifyPolling(activeSid);
+    if(typeof _fetchYoloState==='function')_fetchYoloState(activeSid);
+    attachLiveStream(activeSid,r.stream_id,[]);
+    if(typeof renderSessionList==='function')void renderSessionList();
+  }catch(e){
+    const err=String((e&&e.message)||e||'Goal command failed');
+    S.messages.push({role:'assistant',content:`**Goal command failed:** ${err}`,_ts:Date.now()/1000,_error:true});
+    renderMessages({preserveScroll:true});
+    showToast(err,3000);
+  }
+}
+
 // ── Busy-input mode commands ──────────────────────────────────────────────
 // These commands let users override the default busy_input_mode setting for a
 // specific message.  They are only meaningful while the agent is running.
diff --git a/static/i18n.js b/static/i18n.js
index ffe70f84..41c97275 100644
--- a/static/i18n.js
+++ b/static/i18n.js
@@ -183,6 +183,7 @@ const LOCALES = {
     theme_set: 'Theme: ',
     no_active_session: 'No active session',
     cmd_queue: 'Queue a message for the next turn',
+    cmd_goal: 'Set or inspect a persistent goal',
     cmd_interrupt: 'Cancel current turn and send a new message',
     cmd_steer: 'Inject a mid-turn correction without interrupting the agent',
     cmd_queue_no_msg: 'Usage: /queue <message>',
@@ -1204,6 +1205,7 @@ const LOCALES = {
     theme_set: 'テーマ: ',
     no_active_session: 'アクティブなセッションがありません',
     cmd_queue: '次のターン用にメッセージをキュー',
+    cmd_goal: '永続ゴールを設定または確認',
     cmd_interrupt: '現在のターンをキャンセルして新規メッセージを送信',
     cmd_steer: 'エージェントを中断せずにターン中に修正を注入',
     cmd_queue_no_msg: '使い方: /queue <メッセージ>',
@@ -2186,6 +2188,7 @@ const LOCALES = {
     theme_set: 'Тема: ',
     no_active_session: 'Нет активной сессии',
     cmd_queue: 'Поставить сообщение в очередь на следующий оборот',
+    cmd_goal: 'Задать или проверить постоянную цель',
     cmd_interrupt: 'Прервать текущий оборот и отправить новое сообщение',
     cmd_steer: 'Направить агента исправлением (переходит к прерыванию)',
     cmd_queue_no_msg: 'Использование: /queue <сообщение>',
@@ -3173,6 +3176,7 @@ const LOCALES = {
     theme_set: 'Tema: ',
     no_active_session: 'No hay ninguna sesión activa',
     cmd_queue: 'Poner mensaje en cola para el siguiente turno',
+    cmd_goal: 'Definir o consultar un objetivo persistente',
     cmd_interrupt: 'Cancelar turno actual y enviar nuevo mensaje',
     cmd_steer: 'Inyectar una corrección a mitad del turno sin interrumpir al agente',
     cmd_queue_no_msg: 'Uso: /queue <mensaje>',
@@ -4109,6 +4113,7 @@ const LOCALES = {
     model_scope_advisory: 'Gilt für diesen Chat ab Ihrer nächsten Nachricht.',
     model_scope_toast: 'Gilt für diesen Chat ab Ihrer nächsten Nachricht.',
     cmd_queue: 'Nachricht f\u00fcr den n\u00e4chsten Durchgang einreihen',
+    cmd_goal: 'Ein dauerhaftes Ziel setzen oder prüfen',
     cmd_interrupt: 'Aktuellen Durchgang abbrechen und neue Nachricht senden',
     cmd_steer: 'Korrektursignal einf\u00fcgen ohne Unterbrechung',
     cmd_queue_no_msg: 'Verwendung: /queue <Nachricht>',
@@ -5082,6 +5087,7 @@ const LOCALES = {
     theme_set: '\u4e3b\u9898\uff1a',
     no_active_session: '\u5f53\u524d\u6ca1\u6709\u6d3b\u52a8\u4f1a\u8bdd',
     cmd_queue: '\u5c06\u6d88\u606f\u52a0\u5165\u4e0b\u4e00\u8f6e\u7684\u961f\u5217',
+    cmd_goal: '设置或查看持久目标',
     cmd_interrupt: '\u53d6\u6d88\u5f53\u524d\u56de\u5408\u5e76\u53d1\u9001\u65b0\u6d88\u606f',
     cmd_steer: '\u7528\u7ea0\u6b63\u4fe1\u606f\u5f15\u5bfc\u4ee3\u7406\uff08\u56de\u9000\u4e3a\u4e2d\u65ad\uff09',
     cmd_queue_no_msg: '\u7528\u6cd5\uff1a/queue <\u6d88\u606f>',
@@ -6459,6 +6465,7 @@ const LOCALES = {
     never: '\u5f9e\u4e0d',
     no_active_session: '\u7121\u6d3b\u8e8d\u6703\u8a71',
     cmd_queue: '\u5c07\u8a0a\u606f\u52a0\u5165\u4e0b\u4e00\u8f2a\u7684\u4f47\u5217',
+    cmd_goal: '設定或查看持久目標',
     cmd_interrupt: '\u53d6\u6d88\u7576\u524d\u56de\u5408\u4e26\u767c\u9001\u65b0\u8a0a\u606f',
     cmd_steer: '\u5728\u56de\u5408\u9032\u884c\u4e2d\u6ce8\u5165\u7d3a\u6b63\uff0c\u4e0d\u4e2d\u65b7\u4ee3\u7406',
     cmd_queue_no_msg: '\u7528\u6cd5\uff1a/queue <\u8a0a\u606f>',
@@ -6972,6 +6979,7 @@ const LOCALES = {
     theme_set: 'Tema: ',
     no_active_session: 'Nenhuma sessão ativa',
     cmd_queue: 'Enfileirar mensagem para o próximo turno',
+    cmd_goal: 'Definir ou consultar uma meta persistente',
     cmd_interrupt: 'Cancelar turno atual e enviar nova mensagem',
     cmd_steer: 'Injetar correção no meio do turno sem interromper',
     cmd_queue_no_msg: 'Uso: /queue <mensagem>',
@@ -7869,6 +7877,7 @@ const LOCALES = {
     theme_set: 'Theme: ',
     no_active_session: '활성 세션 없음',
     cmd_queue: 'Queue a message for the next turn',
+    cmd_goal: '지속 목표를 설정하거나 확인',
     cmd_interrupt: 'Cancel current turn and send a new message',
     cmd_steer: 'Inject a mid-turn correction without interrupting the agent',
     cmd_queue_no_msg: 'Usage: /queue <message>',
diff --git a/static/messages.js b/static/messages.js
index 95ca37d5..3c196937 100644
--- a/static/messages.js
+++ b/static/messages.js
@@ -63,7 +63,7 @@ async function send(){
       // cmdSteer / cmdInterrupt say "No active task to stop."
       if(text.startsWith('/')){
         const _pc=typeof parseCommand==='function'&&parseCommand(text);
-        if(_pc&&['steer','interrupt','queue','terminal'].includes(_pc.name)){
+        if(_pc&&['steer','interrupt','queue','terminal','goal'].includes(_pc.name)){
           const _bc=COMMANDS.find(c=>c.name===_pc.name);
           if(_bc){
             $('msg').value='';autoResize();
@@ -336,6 +336,8 @@ function attachLiveStream(activeSid, streamId, uploaded=[], options={}){
   let assistantText='';
   let reasoningText='';
   let liveReasoningText='';
+  let _latestGoalStatus=null;
+  let _pendingGoalContinuation=null;
   let assistantRow=null;
   let assistantBody=null;
   let segmentStart=0;      // char offset in assistantText where current segment begins
@@ -879,6 +881,41 @@ function attachLiveStream(activeSid, streamId, uploaded=[], options={}){
       }catch(_){}
     });
 
+    source.addEventListener('goal',e=>{
+      try{
+        const d=JSON.parse(e.data||'{}');
+        if((d.session_id||activeSid)!==activeSid) return;
+        const goalState=String(d.state||'').trim();
+        const goalEvaluatingMessage='Evaluating goal progress…';
+        if(goalState==='evaluating'){
+          setComposerStatus(goalEvaluatingMessage);
+          return;
+        }
+        const msg=String(d.message||'').trim();
+        if(!msg)return;
+        _latestGoalStatus={message:msg,decision:d.decision||null,state:goalState||null};
+        setComposerStatus(msg);
+        showToast(msg.split('\n')[0],2600);
+      }catch(_){}
+    });
+
+    source.addEventListener('goal_continue',e=>{
+      try{
+        const d=JSON.parse(e.data||'{}');
+        const sid=d.session_id||activeSid;
+        const continuation_prompt=String(d.continuation_prompt||d.text||'').trim();
+        if(!continuation_prompt||sid!==activeSid)return;
+        _pendingGoalContinuation={
+          sid,
+          text:continuation_prompt,
+          model:S.session&&S.session.model||'',
+          model_provider:S.session&&S.session.model_provider||null,
+          profile:S.activeProfile||'default',
+        };
+        showToast('Continuing toward goal…',2200);
+      }catch(_){}
+    });
+
     source.addEventListener('done',e=>{
       _terminalStateReached=true;
       if(_persistTimer){clearTimeout(_persistTimer);_persistTimer=null;}
@@ -992,6 +1029,15 @@ function attachLiveStream(activeSid, streamId, uploaded=[], options={}){
           const lastUser=[...S.messages].reverse().find(m=>m.role==='user');
           if(lastUser)lastUser.attachments=uploaded;
         }
+        if(_latestGoalStatus&&_latestGoalStatus.message){
+          S.messages.push({
+            role:'assistant',
+            content:String(_latestGoalStatus.message),
+            _ts:Date.now()/1000,
+            _goalStatus:true,
+            _transient:true,
+          });
+        }
         clearLiveToolCards();
         S.busy=false;
         // No-reply guard (#373): if agent returned nothing, show inline error
@@ -1003,6 +1049,18 @@ function attachLiveStream(activeSid, streamId, uploaded=[], options={}){
         // TTS auto-read: speak the last assistant response if enabled (#499)
         if(typeof autoReadLastAssistant==='function') setTimeout(()=>autoReadLastAssistant(), 300);
       }
+      if(isActiveSession&&_pendingGoalContinuation&&typeof queueSessionMessage==='function'){
+        const _goalNext=_pendingGoalContinuation;
+        _pendingGoalContinuation=null;
+        queueSessionMessage(_goalNext.sid,{
+          text:_goalNext.text,
+          files:[],
+          model:_goalNext.model,
+          model_provider:_goalNext.model_provider,
+          profile:_goalNext.profile,
+        });
+        if(typeof updateQueueBadge==='function')updateQueueBadge(_goalNext.sid);
+      }
       if(isActiveSession) _queueDrainSid=activeSid;
       renderSessionList();
       _setActivePaneIdleIfOwner();
diff --git a/tests/test_goal_command_webui.py b/tests/test_goal_command_webui.py
new file mode 100644
index 00000000..e1bca59b
--- /dev/null
+++ b/tests/test_goal_command_webui.py
@@ -0,0 +1,272 @@
+"""Regression tests for first-class WebUI /goal command parity."""
+
+import io
+import json
+from pathlib import Path
+from types import SimpleNamespace
+
+import pytest
+
+REPO_ROOT = Path(__file__).resolve().parents[1]
+COMMANDS_JS = (REPO_ROOT / "static" / "commands.js").read_text(encoding="utf-8")
+MESSAGES_JS = (REPO_ROOT / "static" / "messages.js").read_text(encoding="utf-8")
+ROUTES_PY = (REPO_ROOT / "api" / "routes.py").read_text(encoding="utf-8")
+STREAMING_PY = (REPO_ROOT / "api" / "streaming.py").read_text(encoding="utf-8")
+
+
+def test_goal_command_payload_matches_gateway_controls(monkeypatch):
+    """The backend command helper mirrors gateway /goal status/pause/resume/clear/set."""
+    from api import goals as webui_goals
+
+    calls = []
+
+    class FakeState:
+        goal = "ship the feature"
+        status = "active"
+        turns_used = 0
+        max_turns = 20
+        last_verdict = None
+        last_reason = None
+        paused_reason = None
+
+    class FakeGoalManager:
+        def __init__(self, session_id, default_max_turns=20):
+            calls.append(("init", session_id, default_max_turns))
+            self.state = None
+
+        def status_line(self):
+            return "No active goal. Set one with /goal <text>."
+
+        def pause(self, reason="user-paused"):
+            calls.append(("pause", reason))
+            return FakeState()
+
+        def resume(self, reset_budget=True):
+            calls.append(("resume", reset_budget))
+            return FakeState()
+
+        def has_goal(self):
+            return True
+
+        def clear(self):
+            calls.append(("clear",))
+
+        def set(self, goal):
+            calls.append(("set", goal))
+            state = FakeState()
+            state.goal = goal
+            self.state = state
+            return state
+
+    monkeypatch.setattr(webui_goals, "GoalManager", FakeGoalManager)
+    monkeypatch.setattr(webui_goals, "_default_max_turns", lambda: 20)
+
+    status = webui_goals.goal_command_payload("sid-123", "status")
+    pause = webui_goals.goal_command_payload("sid-123", "pause")
+    resume = webui_goals.goal_command_payload("sid-123", "resume")
+    clear = webui_goals.goal_command_payload("sid-123", "clear")
+    set_goal = webui_goals.goal_command_payload("sid-123", "ship the feature")
+
+    assert status["message"] == "No active goal. Set one with /goal <text>."
+    assert pause["message"] == "⏸ Goal paused: ship the feature"
+    assert resume["message"].startswith("▶ Goal resumed: ship the feature")
+    assert clear["message"] == "Goal cleared."
+    assert set_goal["action"] == "set"
+    assert set_goal["kickoff_prompt"] == "ship the feature"
+    assert "⊙ Goal set (20-turn budget): ship the feature" in set_goal["message"]
+    assert ("set", "ship the feature") in calls
+
+
+def test_goal_command_payload_rejects_new_goal_while_stream_running(monkeypatch):
+    """Status/control subcommands are safe mid-run; replacing the goal is not."""
+    from api import goals as webui_goals
+
+    class FakeGoalManager:
+        def __init__(self, session_id, default_max_turns=20):
+            pass
+
+        def status_line(self):
+            return "⊙ Goal (active, 1/20 turns): existing"
+
+    monkeypatch.setattr(webui_goals, "GoalManager", FakeGoalManager)
+    monkeypatch.setattr(webui_goals, "_default_max_turns", lambda: 20)
+
+    status = webui_goals.goal_command_payload("sid-123", "status", stream_running=True)
+    rejected = webui_goals.goal_command_payload("sid-123", "replace it", stream_running=True)
+
+    assert status["ok"] is True
+    assert rejected["ok"] is False
+    assert rejected["error"] == "agent_running"
+    assert "use /goal status / pause / clear mid-run" in rejected["message"]
+
+
+def test_has_active_goal_reports_only_active_state(monkeypatch):
+    """Streaming can avoid showing an evaluating spinner when no standing goal is active."""
+    from api import goals as webui_goals
+
+    class FakeGoalManager:
+        def __init__(self, session_id, default_max_turns=20):
+            self.session_id = session_id
+
+        def is_active(self):
+            return self.session_id == "sid-active-goal"
+
+    monkeypatch.setattr(webui_goals, "GoalManager", FakeGoalManager)
+    monkeypatch.setattr(webui_goals, "_default_max_turns", lambda: 20)
+
+    assert webui_goals.has_active_goal("sid-active-goal") is True
+    assert webui_goals.has_active_goal("sid-idle-goal") is False
+    assert webui_goals.has_active_goal("") is False
+
+
+def test_goal_continuation_decision_emits_status_and_normal_user_prompt(monkeypatch):
+    """Post-turn hook returns the visible status event plus a normal continuation prompt."""
+    from api import goals as webui_goals
+
+    class FakeGoalManager:
+        def __init__(self, session_id, default_max_turns=20):
+            self.session_id = session_id
+
+        def is_active(self):
+            return True
+
+        def evaluate_after_turn(self, last_response, user_initiated=True):
+            return {
+                "status": "active",
+                "should_continue": True,
+                "continuation_prompt": "[Continuing toward your standing goal]\nGoal: ship it",
+                "verdict": "continue",
+                "reason": "one step remains",
+                "message": "↻ Continuing toward goal (1/20): one step remains",
+            }
+
+    monkeypatch.setattr(webui_goals, "GoalManager", FakeGoalManager)
+    monkeypatch.setattr(webui_goals, "_default_max_turns", lambda: 20)
+
+    decision = webui_goals.evaluate_goal_after_turn("sid-123", "not done yet", user_initiated=False)
+
+    assert decision["message"].startswith("↻ Continuing toward goal")
+    assert decision["should_continue"] is True
+    assert decision["continuation_prompt"].startswith("[Continuing toward your standing goal]")
+
+
+def test_goal_endpoint_sets_goal_and_starts_kickoff_stream(monkeypatch, tmp_path):
+    """POST /api/goal uses GoalManager state and launches the first goal turn."""
+    from api import goals as webui_goals
+    from api import routes
+
+    class FakeState:
+        goal = "ship the feature"
+        status = "active"
+        turns_used = 0
+        max_turns = 20
+        last_verdict = None
+        last_reason = None
+        paused_reason = None
+
+    class FakeGoalManager:
+        def __init__(self, session_id, default_max_turns=20):
+            self.session_id = session_id
+            self.default_max_turns = default_max_turns
+
+        def set(self, goal):
+            state = FakeState()
+            state.goal = goal
+            return state
+
+    class FakeSession:
+        session_id = "sid-goal-route"
+        profile = "default"
+        workspace = str(tmp_path)
+        model = "gpt-5.5"
+        model_provider = "openai-codex"
+        messages = []
+        context_messages = []
+        pending_user_message = None
+        active_stream_id = None
+
+    monkeypatch.setattr(webui_goals, "GoalManager", FakeGoalManager)
+    monkeypatch.setattr(routes, "get_session", lambda sid: FakeSession())
+    monkeypatch.setattr(routes, "resolve_trusted_workspace", lambda workspace: tmp_path)
+    monkeypatch.setattr(
+        routes,
+        "_resolve_compatible_session_model_state",
+        lambda model, provider: (model, provider, False),
+    )
+    started = []
+
+    def fake_start(session, **kwargs):
+        started.append(kwargs)
+        return {"stream_id": "goal-stream", "session_id": session.session_id, "pending_started_at": 123.0}
+
+    monkeypatch.setattr(routes, "_start_chat_stream_for_session", fake_start)
+    monkeypatch.setattr(routes, "j", lambda handler, payload, status=200, **kwargs: {"status": status, "payload": payload})
+
+    result = routes._handle_goal_command(
+        object(),
+        {
+            "session_id": "sid-goal-route",
+            "args": "ship the feature",
+            "workspace": str(tmp_path),
+            "model": "gpt-5.5",
+            "model_provider": "openai-codex",
+        },
+    )
+
+    assert result["status"] == 200
+    assert result["payload"]["action"] == "set"
+    assert result["payload"]["stream_id"] == "goal-stream"
+    assert started and started[0]["msg"] == "ship the feature"
+    assert started[0]["model_provider"] == "openai-codex"
+
+
+def test_routes_register_goal_endpoint_and_kickoff_stream():
+    assert 'if parsed.path == "/api/goal"' in ROUTES_PY
+    assert "return _handle_goal_command(handler, body)" in ROUTES_PY
+    assert "goal_command_payload" in ROUTES_PY
+    assert "kickoff_prompt" in ROUTES_PY
+    assert "_start_chat_stream_for_session" in ROUTES_PY
+
+
+def test_streaming_post_turn_goal_hook_surfaces_and_continues():
+    assert "evaluate_goal_after_turn" in STREAMING_PY
+    assert "put('goal'" in STREAMING_PY
+    assert "decision.get('should_continue')" in STREAMING_PY
+    assert "continuation_prompt" in STREAMING_PY
+    assert "put('goal_continue'" in STREAMING_PY
+    goal_idx = STREAMING_PY.find("evaluate_goal_after_turn")
+    done_idx = STREAMING_PY.find("put('done'", goal_idx)
+    assert goal_idx != -1 and done_idx != -1
+    assert goal_idx < done_idx, "goal status should be emitted before the terminal done payload"
+
+
+def test_streaming_goal_hook_emits_evaluating_state_before_judge():
+    evaluating_idx = STREAMING_PY.find("'state': 'evaluating'")
+    judge_idx = STREAMING_PY.find("_goal_decision = evaluate_goal_after_turn")
+    done_idx = STREAMING_PY.find("put('done'", judge_idx)
+    assert evaluating_idx != -1, "goal hook should emit an evaluating state before judge round-trip"
+    assert judge_idx != -1 and done_idx != -1
+    assert evaluating_idx < judge_idx < done_idx
+    assert "Evaluating goal progress…" in STREAMING_PY
+    assert "'state': 'continuing' if decision.get('should_continue') else 'idle'" in STREAMING_PY
+
+
+def test_frontend_has_goal_slash_command_and_status_event_handler():
+    assert "{name:'goal'" in COMMANDS_JS
+    assert "subArgs:['status','pause','resume','clear']" in COMMANDS_JS
+    assert "function cmdGoal" in COMMANDS_JS
+    assert "api('/api/goal'" in COMMANDS_JS
+    assert "stream_id" in COMMANDS_JS
+    assert "goal'" in MESSAGES_JS
+    assert "source.addEventListener('goal'" in MESSAGES_JS
+    assert "source.addEventListener('goal_continue'" in MESSAGES_JS
+    assert "['steer','interrupt','queue','terminal','goal'].includes(_pc.name)" in MESSAGES_JS
+    assert "queueSessionMessage" in MESSAGES_JS
+
+
+def test_frontend_goal_evaluating_state_uses_calm_composer_indicator():
+    assert "const goalState=String(d.state||'').trim();" in MESSAGES_JS
+    assert "const goalEvaluatingMessage='Evaluating goal progress…';" in MESSAGES_JS
+    assert "if(goalState==='evaluating')" in MESSAGES_JS
+    assert "setComposerStatus(goalEvaluatingMessage);" in MESSAGES_JS
+    assert "return;" in MESSAGES_JS
diff --git a/tests/test_mcp_server.py b/tests/test_mcp_server.py
index 7e0315ec..5322d1d7 100644
--- a/tests/test_mcp_server.py
+++ b/tests/test_mcp_server.py
@@ -49,26 +49,67 @@ def _cleanup_state_dir(state_dir: Path):
 
 
 def _reimport_mcp():
-    """Re-import mcp_server with current env vars and profile.
+    """Re-point mcp_server's module-level STATE_DIR / SESSION_DIR /
+    SESSION_INDEX_FILE / PROJECTS_FILE constants at the current
+    HERMES_WEBUI_STATE_DIR.
 
     Returns (mcp_module, profiles_module) — profiles_module is the
-    live api.profiles reference that the re-imported mcp_server uses.
+    live api.profiles reference.
+
+    NOTE: Does NOT use `del sys.modules[...]` or `importlib.reload(...)`.
+    Both patterns trigger a chain re-import inside the FastMCP / pydantic
+    stack that corrupts pydantic's `_generics._GENERIC_TYPES_CACHE`
+    (manifests as `KeyError: 'pydantic.root_model'` in unrelated
+    downstream tests in the full suite). Instead, we mutate the
+    constants in-place after the first one-time import, which is
+    behaviorally equivalent for these tests since the constants are
+    module-level Path objects used only to compute STATE_DIR-rooted
+    paths at call time.
     """
-    # Clear cached module and api submodules that cache paths
-    for key in list(sys.modules.keys()):
-        if key == 'mcp_server' or key.startswith('mcp_server.') or \
-           key == 'api.config' or key == 'api.models' or key == 'api.profiles':
-            del sys.modules[key]
+    state_dir = Path(os.environ['HERMES_WEBUI_STATE_DIR'])
 
-    import importlib
     import api.config as cfg
-    importlib.reload(cfg)
-
-    # Re-acquire api.profiles reference (old one is stale after sys.modules clear)
     import api.profiles as fresh_profiles
-    fresh_profiles._active_profile = 'default'
-
     import mcp_server as mod
+
+    # Re-point api.config module-level constants
+    cfg.STATE_DIR = state_dir
+    cfg.SESSION_DIR = state_dir / "sessions"
+    cfg.WORKSPACES_FILE = state_dir / "workspaces.json"
+    cfg.SETTINGS_FILE = state_dir / "settings.json"
+    cfg.LAST_WORKSPACE_FILE = state_dir / "last_workspace.txt"
+    cfg.PROJECTS_FILE = state_dir / "projects.json"
+    if hasattr(cfg, 'SESSION_INDEX_FILE'):
+        cfg.SESSION_INDEX_FILE = state_dir / "sessions" / "_index.json"
+
+    # Re-point mcp_server's imported aliases (they were copied at first
+    # import and don't pick up cfg mutations automatically).
+    mod.STATE_DIR = cfg.STATE_DIR
+    mod.SESSION_DIR = cfg.SESSION_DIR
+    mod.PROJECTS_FILE = cfg.PROJECTS_FILE
+    if hasattr(mod, 'SESSION_INDEX_FILE'):
+        mod.SESSION_INDEX_FILE = cfg.SESSION_INDEX_FILE
+
+    # api.models also imports STATE_DIR / PROJECTS_FILE etc. as module
+    # constants — re-point those too so load_projects() / save_projects()
+    # see the fresh STATE_DIR.
+    if 'api.models' in sys.modules:
+        models_mod = sys.modules['api.models']
+        if hasattr(models_mod, 'STATE_DIR'):
+            models_mod.STATE_DIR = cfg.STATE_DIR
+        if hasattr(models_mod, 'PROJECTS_FILE'):
+            models_mod.PROJECTS_FILE = cfg.PROJECTS_FILE
+        if hasattr(models_mod, 'SESSION_DIR'):
+            models_mod.SESSION_DIR = cfg.SESSION_DIR
+
+    # Re-evaluate WEBUI_URL from current env (PR #1895 made it env-aware
+    # but the value is computed once at module load; tests need to see
+    # current env state).
+    mod.WEBUI_HOST = os.environ.get("HERMES_WEBUI_HOST", "127.0.0.1")
+    mod.WEBUI_PORT = os.environ.get("HERMES_WEBUI_PORT", "8787")
+    mod.WEBUI_URL = f"http://{mod.WEBUI_HOST}:{mod.WEBUI_PORT}"
+
+    fresh_profiles._active_profile = 'default'
     return mod, fresh_profiles
 
 
@@ -292,7 +333,8 @@ class TestProfileScoping:
         to 'default'. A non-root profile must NOT see legacy untagged rows.
         """
         # Manually write a legacy untagged project (pre-#1614 schema)
-        from api.config import PROJECTS_FILE
+        import api.config as _cfg_mod
+        PROJECTS_FILE = _cfg_mod.PROJECTS_FILE
         legacy = [{
             "project_id": "legacy000001",
             "name": "LegacyUntagged",
@@ -317,7 +359,8 @@ class TestProfileScoping:
 
     async def test_legacy_untagged_rename_blocked_from_non_root(self):
         """Non-root profile cannot rename a legacy untagged project."""
-        from api.config import PROJECTS_FILE
+        import api.config as _cfg_mod
+        PROJECTS_FILE = _cfg_mod.PROJECTS_FILE
         legacy = [{
             "project_id": "legacy000002",
             "name": "Legacy",
@@ -517,7 +560,8 @@ class TestProfileCliOrdering:
         _profiles._active_profile = _profile_arg right after import). If a
         helper had latched the profile at import time, the override here
         would be too late and the test would see 'default'-tagged rows."""
-        from api.config import PROJECTS_FILE
+        import api.config as _cfg_mod
+        PROJECTS_FILE = _cfg_mod.PROJECTS_FILE
         # Pre-seed two projects: one for default, one for foo.
         seeded = [
             {"project_id": "p_default_0001", "name": "DefaultRow",