fix: restore /usage account limits (#958)

agent/account_usage.py

@@ -0,0 +1,326 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+from datetime import datetime, timezone
+from typing import Any, Optional
+
+import httpx
+
+from agent.anthropic_adapter import _is_oauth_token, resolve_anthropic_token
+from hermes_cli.auth import _read_codex_tokens, resolve_codex_runtime_credentials
+from hermes_cli.runtime_provider import resolve_runtime_provider
+
+
+def _utc_now() -> datetime:
+    return datetime.now(timezone.utc)
+
+
+@dataclass(frozen=True)
+class AccountUsageWindow:
+    label: str
+    used_percent: Optional[float] = None
+    reset_at: Optional[datetime] = None
+    detail: Optional[str] = None
+
+
+@dataclass(frozen=True)
+class AccountUsageSnapshot:
+    provider: str
+    source: str
+    fetched_at: datetime
+    title: str = "Account limits"
+    plan: Optional[str] = None
+    windows: tuple[AccountUsageWindow, ...] = ()
+    details: tuple[str, ...] = ()
+    unavailable_reason: Optional[str] = None
+
+    @property
+    def available(self) -> bool:
+        return bool(self.windows or self.details) and not self.unavailable_reason
+
+
+def _title_case_slug(value: Optional[str]) -> Optional[str]:
+    cleaned = str(value or "").strip()
+    if not cleaned:
+        return None
+    return cleaned.replace("_", " ").replace("-", " ").title()
+
+
+def _parse_dt(value: Any) -> Optional[datetime]:
+    if value in (None, ""):
+        return None
+    if isinstance(value, (int, float)):
+        return datetime.fromtimestamp(float(value), tz=timezone.utc)
+    if isinstance(value, str):
+        text = value.strip()
+        if not text:
+            return None
+        if text.endswith("Z"):
+            text = text[:-1] + "+00:00"
+        try:
+            dt = datetime.fromisoformat(text)
+            return dt if dt.tzinfo else dt.replace(tzinfo=timezone.utc)
+        except ValueError:
+            return None
+    return None
+
+
+def _format_reset(dt: Optional[datetime]) -> str:
+    if not dt:
+        return "unknown"
+    local_dt = dt.astimezone()
+    delta = dt - _utc_now()
+    total_seconds = int(delta.total_seconds())
+    if total_seconds <= 0:
+        return f"now ({local_dt.strftime('%Y-%m-%d %H:%M %Z')})"
+    hours, rem = divmod(total_seconds, 3600)
+    minutes = rem // 60
+    if hours >= 24:
+        days, hours = divmod(hours, 24)
+        rel = f"in {days}d {hours}h"
+    elif hours > 0:
+        rel = f"in {hours}h {minutes}m"
+    else:
+        rel = f"in {minutes}m"
+    return f"{rel} ({local_dt.strftime('%Y-%m-%d %H:%M %Z')})"
+
+
+def render_account_usage_lines(snapshot: Optional[AccountUsageSnapshot], *, markdown: bool = False) -> list[str]:
+    if not snapshot:
+        return []
+    header = f"📈 {'**' if markdown else ''}{snapshot.title}{'**' if markdown else ''}"
+    lines = [header]
+    if snapshot.plan:
+        lines.append(f"Provider: {snapshot.provider} ({snapshot.plan})")
+    else:
+        lines.append(f"Provider: {snapshot.provider}")
+    for window in snapshot.windows:
+        if window.used_percent is None:
+            base = f"{window.label}: unavailable"
+        else:
+            remaining = max(0, round(100 - float(window.used_percent)))
+            used = max(0, round(float(window.used_percent)))
+            base = f"{window.label}: {remaining}% remaining ({used}% used)"
+        if window.reset_at:
+            base += f" • resets {_format_reset(window.reset_at)}"
+        elif window.detail:
+            base += f" • {window.detail}"
+        lines.append(base)
+    for detail in snapshot.details:
+        lines.append(detail)
+    if snapshot.unavailable_reason:
+        lines.append(f"Unavailable: {snapshot.unavailable_reason}")
+    return lines
+
+
+def _resolve_codex_usage_url(base_url: str) -> str:
+    normalized = (base_url or "").strip().rstrip("/")
+    if not normalized:
+        normalized = "https://chatgpt.com/backend-api/codex"
+    if normalized.endswith("/codex"):
+        normalized = normalized[: -len("/codex")]
+    if "/backend-api" in normalized:
+        return normalized + "/wham/usage"
+    return normalized + "/api/codex/usage"
+
+
+def _fetch_codex_account_usage() -> Optional[AccountUsageSnapshot]:
+    creds = resolve_codex_runtime_credentials(refresh_if_expiring=True)
+    token_data = _read_codex_tokens()
+    tokens = token_data.get("tokens") or {}
+    account_id = str(tokens.get("account_id", "") or "").strip() or None
+    headers = {
+        "Authorization": f"Bearer {creds['api_key']}",
+        "Accept": "application/json",
+        "User-Agent": "codex-cli",
+    }
+    if account_id:
+        headers["ChatGPT-Account-Id"] = account_id
+    with httpx.Client(timeout=15.0) as client:
+        response = client.get(_resolve_codex_usage_url(creds.get("base_url", "")), headers=headers)
+        response.raise_for_status()
+    payload = response.json() or {}
+    rate_limit = payload.get("rate_limit") or {}
+    windows: list[AccountUsageWindow] = []
+    for key, label in (("primary_window", "Session"), ("secondary_window", "Weekly")):
+        window = rate_limit.get(key) or {}
+        used = window.get("used_percent")
+        if used is None:
+            continue
+        windows.append(
+            AccountUsageWindow(
+                label=label,
+                used_percent=float(used),
+                reset_at=_parse_dt(window.get("reset_at")),
+            )
+        )
+    details: list[str] = []
+    credits = payload.get("credits") or {}
+    if credits.get("has_credits"):
+        balance = credits.get("balance")
+        if isinstance(balance, (int, float)):
+            details.append(f"Credits balance: ${float(balance):.2f}")
+        elif credits.get("unlimited"):
+            details.append("Credits balance: unlimited")
+    return AccountUsageSnapshot(
+        provider="openai-codex",
+        source="usage_api",
+        fetched_at=_utc_now(),
+        plan=_title_case_slug(payload.get("plan_type")),
+        windows=tuple(windows),
+        details=tuple(details),
+    )
+
+
+def _fetch_anthropic_account_usage() -> Optional[AccountUsageSnapshot]:
+    token = (resolve_anthropic_token() or "").strip()
+    if not token:
+        return None
+    if not _is_oauth_token(token):
+        return AccountUsageSnapshot(
+            provider="anthropic",
+            source="oauth_usage_api",
+            fetched_at=_utc_now(),
+            unavailable_reason="Anthropic account limits are only available for OAuth-backed Claude accounts.",
+        )
+    headers = {
+        "Authorization": f"Bearer {token}",
+        "Accept": "application/json",
+        "Content-Type": "application/json",
+        "anthropic-beta": "oauth-2025-04-20",
+        "User-Agent": "claude-code/2.1.0",
+    }
+    with httpx.Client(timeout=15.0) as client:
+        response = client.get("https://api.anthropic.com/api/oauth/usage", headers=headers)
+        response.raise_for_status()
+    payload = response.json() or {}
+    windows: list[AccountUsageWindow] = []
+    mapping = (
+        ("five_hour", "Current session"),
+        ("seven_day", "Current week"),
+        ("seven_day_opus", "Opus week"),
+        ("seven_day_sonnet", "Sonnet week"),
+    )
+    for key, label in mapping:
+        window = payload.get(key) or {}
+        util = window.get("utilization")
+        if util is None:
+            continue
+        used = float(util) * 100 if float(util) <= 1 else float(util)
+        windows.append(
+            AccountUsageWindow(
+                label=label,
+                used_percent=used,
+                reset_at=_parse_dt(window.get("resets_at")),
+            )
+        )
+    details: list[str] = []
+    extra = payload.get("extra_usage") or {}
+    if extra.get("is_enabled"):
+        used_credits = extra.get("used_credits")
+        monthly_limit = extra.get("monthly_limit")
+        currency = extra.get("currency") or "USD"
+        if isinstance(used_credits, (int, float)) and isinstance(monthly_limit, (int, float)):
+            details.append(
+                f"Extra usage: {used_credits:.2f} / {monthly_limit:.2f} {currency}"
+            )
+    return AccountUsageSnapshot(
+        provider="anthropic",
+        source="oauth_usage_api",
+        fetched_at=_utc_now(),
+        windows=tuple(windows),
+        details=tuple(details),
+    )
+
+
+def _fetch_openrouter_account_usage(base_url: Optional[str], api_key: Optional[str]) -> Optional[AccountUsageSnapshot]:
+    runtime = resolve_runtime_provider(
+        requested="openrouter",
+        explicit_base_url=base_url,
+        explicit_api_key=api_key,
+    )
+    token = str(runtime.get("api_key", "") or "").strip()
+    if not token:
+        return None
+    normalized = str(runtime.get("base_url", "") or "").rstrip("/")
+    credits_url = f"{normalized}/credits"
+    key_url = f"{normalized}/key"
+    headers = {
+        "Authorization": f"Bearer {token}",
+        "Accept": "application/json",
+    }
+    with httpx.Client(timeout=10.0) as client:
+        credits_resp = client.get(credits_url, headers=headers)
+        credits_resp.raise_for_status()
+        credits = (credits_resp.json() or {}).get("data") or {}
+        try:
+            key_resp = client.get(key_url, headers=headers)
+            key_resp.raise_for_status()
+            key_data = (key_resp.json() or {}).get("data") or {}
+        except Exception:
+            key_data = {}
+    total_credits = float(credits.get("total_credits") or 0.0)
+    total_usage = float(credits.get("total_usage") or 0.0)
+    details = [f"Credits balance: ${max(0.0, total_credits - total_usage):.2f}"]
+    windows: list[AccountUsageWindow] = []
+    limit = key_data.get("limit")
+    limit_remaining = key_data.get("limit_remaining")
+    limit_reset = str(key_data.get("limit_reset") or "").strip()
+    usage = key_data.get("usage")
+    if (
+        isinstance(limit, (int, float))
+        and float(limit) > 0
+        and isinstance(limit_remaining, (int, float))
+        and 0 <= float(limit_remaining) <= float(limit)
+    ):
+        limit_value = float(limit)
+        remaining_value = float(limit_remaining)
+        used_percent = ((limit_value - remaining_value) / limit_value) * 100
+        detail_parts = [f"${remaining_value:.2f} of ${limit_value:.2f} remaining"]
+        if limit_reset:
+            detail_parts.append(f"resets {limit_reset}")
+        windows.append(
+            AccountUsageWindow(
+                label="API key quota",
+                used_percent=used_percent,
+                detail=" • ".join(detail_parts),
+            )
+        )
+    if isinstance(usage, (int, float)):
+        usage_parts = [f"API key usage: ${float(usage):.2f} total"]
+        for value, label in (
+            (key_data.get("usage_daily"), "today"),
+            (key_data.get("usage_weekly"), "this week"),
+            (key_data.get("usage_monthly"), "this month"),
+        ):
+            if isinstance(value, (int, float)) and float(value) > 0:
+                usage_parts.append(f"${float(value):.2f} {label}")
+        details.append(" • ".join(usage_parts))
+    return AccountUsageSnapshot(
+        provider="openrouter",
+        source="credits_api",
+        fetched_at=_utc_now(),
+        windows=tuple(windows),
+        details=tuple(details),
+    )
+
+
+def fetch_account_usage(
+    provider: Optional[str],
+    *,
+    base_url: Optional[str] = None,
+    api_key: Optional[str] = None,
+) -> Optional[AccountUsageSnapshot]:
+    normalized = str(provider or "").strip().lower()
+    if normalized in {"", "auto", "custom"}:
+        return None
+    try:
+        if normalized == "openai-codex":
+            return _fetch_codex_account_usage()
+        if normalized == "anthropic":
+            return _fetch_anthropic_account_usage()
+        if normalized == "openrouter":
+            return _fetch_openrouter_account_usage(base_url, api_key)
+    except Exception:
+        return None
+    return None

cli.py

@@ -13,6 +13,7 @@ Usage:
     python cli.py --list-tools             # List available tools and exit
 """
 
+import concurrent.futures
 import logging
 import os
 import shutil
@@ -63,6 +64,7 @@ from agent.usage_pricing import (
     format_duration_compact,
     format_token_count_compact,
 )
+from agent.account_usage import fetch_account_usage, render_account_usage_lines
 from hermes_cli.banner import _format_context_length, format_banner_version_label
 
 _COMMAND_SPINNER_FRAMES = ("⠋", "⠙", "⠹", "⠸", "⠼", "⠴", "⠦", "⠧", "⠇", "⠏")
@@ -6471,6 +6473,29 @@ class HermesCLI:
         if cost_result.status == "unknown":
             print(f"  Note:             Pricing unknown for {agent.model}")
 
+        # Account limits -- fetched off-thread with a hard timeout so slow
+        # provider APIs don't hang the prompt.
+        provider = getattr(agent, "provider", None) or getattr(self, "provider", None)
+        base_url = getattr(agent, "base_url", None) or getattr(self, "base_url", None)
+        api_key = getattr(agent, "api_key", None) or getattr(self, "api_key", None)
+        account_snapshot = None
+        if provider:
+            with concurrent.futures.ThreadPoolExecutor(max_workers=1) as _pool:
+                try:
+                    account_snapshot = _pool.submit(
+                        fetch_account_usage,
+                        provider,
+                        base_url=base_url,
+                        api_key=api_key,
+                    ).result(timeout=10.0)
+                except (concurrent.futures.TimeoutError, Exception):
+                    account_snapshot = None
+        account_lines = [f"  {line}" for line in render_account_usage_lines(account_snapshot)]
+        if account_lines:
+            print()
+            for line in account_lines:
+                print(line)
+
         if self.verbose:
             logging.getLogger().setLevel(logging.DEBUG)
             for noisy in ('openai', 'openai._base_client', 'httpx', 'httpcore', 'asyncio', 'hpack', 'grpc', 'modal'):
@@ -6852,12 +6877,11 @@ class HermesCLI:
             self._voice_stop_and_transcribe()
 
         # Audio cue: single beep BEFORE starting stream (avoid CoreAudio conflict)
-        if self._voice_beeps_enabled():
-            try:
-                from tools.voice_mode import play_beep
-                play_beep(frequency=880, count=1)
-            except Exception:
-                pass
+        try:
+            from tools.voice_mode import play_beep
+            play_beep(frequency=880, count=1)
+        except Exception:
+            pass
 
         try:
             self._voice_recorder.start(on_silence_stop=_on_silence)
@@ -6905,12 +6929,11 @@ class HermesCLI:
             wav_path = self._voice_recorder.stop()
 
             # Audio cue: double beep after stream stopped (no CoreAudio conflict)
-            if self._voice_beeps_enabled():
-                try:
-                    from tools.voice_mode import play_beep
-                    play_beep(frequency=660, count=2)
-                except Exception:
-                    pass
+            try:
+                from tools.voice_mode import play_beep
+                play_beep(frequency=660, count=2)
+            except Exception:
+                pass
 
             if wav_path is None:
                 _cprint(f"{_DIM}No speech detected.{_RST}")
@@ -7061,17 +7084,6 @@ class HermesCLI:
             _cprint(f"Unknown voice subcommand: {subcommand}")
             _cprint("Usage: /voice [on|off|tts|status]")
 
-    def _voice_beeps_enabled(self) -> bool:
-        """Return whether CLI voice mode should play record start/stop beeps."""
-        try:
-            from hermes_cli.config import load_config
-            voice_cfg = load_config().get("voice", {})
-            if isinstance(voice_cfg, dict):
-                return bool(voice_cfg.get("beep_enabled", True))
-        except Exception:
-            pass
-        return True
-
     def _enable_voice_mode(self):
         """Enable voice mode after checking requirements."""
         if self._voice_mode:

gateway/run.py

@@ -28,6 +28,8 @@ from pathlib import Path
 from datetime import datetime
 from typing import Dict, Optional, Any, List
 
+from agent.account_usage import fetch_account_usage, render_account_usage_lines
+
 # ---------------------------------------------------------------------------
 # SSL certificate auto-detection for NixOS and other non-standard systems.
 # Must run BEFORE any HTTP library (discord, aiohttp, etc.) is imported.
@@ -6481,6 +6483,38 @@ class GatewayRunner:
                     if cached:
                         agent = cached[0]
 
+        # Resolve provider/base_url/api_key for the account-usage fetch.
+        # Prefer the live agent; fall back to persisted billing data on the
+        # SessionDB row so `/usage` still returns account info between turns
+        # when no agent is resident.
+        provider = getattr(agent, "provider", None) if agent and agent is not _AGENT_PENDING_SENTINEL else None
+        base_url = getattr(agent, "base_url", None) if agent and agent is not _AGENT_PENDING_SENTINEL else None
+        api_key = getattr(agent, "api_key", None) if agent and agent is not _AGENT_PENDING_SENTINEL else None
+        if not provider and getattr(self, "_session_db", None) is not None:
+            try:
+                _entry_for_billing = self.session_store.get_or_create_session(source)
+                persisted = self._session_db.get_session(_entry_for_billing.session_id) or {}
+            except Exception:
+                persisted = {}
+            provider = provider or persisted.get("billing_provider")
+            base_url = base_url or persisted.get("billing_base_url")
+
+        # Fetch account usage off the event loop so slow provider APIs don't
+        # block the gateway. Failures are non-fatal -- account_lines stays [].
+        account_lines: list[str] = []
+        if provider:
+            try:
+                account_snapshot = await asyncio.to_thread(
+                    fetch_account_usage,
+                    provider,
+                    base_url=base_url,
+                    api_key=api_key,
+                )
+            except Exception:
+                account_snapshot = None
+            if account_snapshot:
+                account_lines = render_account_usage_lines(account_snapshot, markdown=True)
+
         if agent and hasattr(agent, "session_total_tokens") and agent.session_api_calls > 0:
             lines = []
 
@@ -6538,6 +6572,10 @@ class GatewayRunner:
             if ctx.compression_count:
                 lines.append(f"Compressions: {ctx.compression_count}")
 
+            if account_lines:
+                lines.append("")
+                lines.extend(account_lines)
+
             return "\n".join(lines)
 
         # No agent at all -- check session history for a rough count
@@ -6547,12 +6585,18 @@ class GatewayRunner:
             from agent.model_metadata import estimate_messages_tokens_rough
             msgs = [m for m in history if m.get("role") in ("user", "assistant") and m.get("content")]
             approx = estimate_messages_tokens_rough(msgs)
-            return (
-                f"📊 **Session Info**\n"
-                f"Messages: {len(msgs)}\n"
-                f"Estimated context: ~{approx:,} tokens\n"
-                f"_(Detailed usage available after the first agent response)_"
-            )
+            lines = [
+                "📊 **Session Info**",
+                f"Messages: {len(msgs)}",
+                f"Estimated context: ~{approx:,} tokens",
+                "_(Detailed usage available after the first agent response)_",
+            ]
+            if account_lines:
+                lines.append("")
+                lines.extend(account_lines)
+            return "\n".join(lines)
+        if account_lines:
+            return "\n".join(account_lines)
         return "No usage data available for this session."
 
     async def _handle_insights_command(self, event: MessageEvent) -> str:

tests/gateway/test_usage_command.py

@@ -175,3 +175,79 @@ class TestUsageCachedAgent:
             result = await runner._handle_usage_command(event)
 
         assert "Cost: included" in result
+
+
+class TestUsageAccountSection:
+    """Account-limits section appended to /usage output."""
+
+    @pytest.mark.asyncio
+    async def test_usage_command_includes_account_section(self, monkeypatch):
+        agent = _make_mock_agent(provider="openai-codex")
+        agent.base_url = "https://chatgpt.com/backend-api/codex"
+        agent.api_key = "unused"
+        runner = _make_runner(SK, cached_agent=agent)
+        event = MagicMock()
+
+        monkeypatch.setattr(
+            "gateway.run.fetch_account_usage",
+            lambda provider, base_url=None, api_key=None: object(),
+        )
+        monkeypatch.setattr(
+            "gateway.run.render_account_usage_lines",
+            lambda snapshot, markdown=False: [
+                "📈 **Account limits**",
+                "Provider: openai-codex (Pro)",
+                "Session: 85% remaining (15% used)",
+            ],
+        )
+        with patch("agent.rate_limit_tracker.format_rate_limit_compact", return_value="RPM: 50/60"), \
+             patch("agent.usage_pricing.estimate_usage_cost") as mock_cost:
+            mock_cost.return_value = MagicMock(amount_usd=None, status="included")
+            result = await runner._handle_usage_command(event)
+
+        assert "📊 **Session Token Usage**" in result
+        assert "📈 **Account limits**" in result
+        assert "Provider: openai-codex (Pro)" in result
+
+    @pytest.mark.asyncio
+    async def test_usage_command_uses_persisted_provider_when_agent_not_running(self, monkeypatch):
+        runner = _make_runner(SK)
+        runner._session_db = MagicMock()
+        runner._session_db.get_session.return_value = {
+            "billing_provider": "openai-codex",
+            "billing_base_url": "https://chatgpt.com/backend-api/codex",
+        }
+        session_entry = MagicMock()
+        session_entry.session_id = "sess-1"
+        runner.session_store.get_or_create_session.return_value = session_entry
+        runner.session_store.load_transcript.return_value = [
+            {"role": "user", "content": "earlier"},
+        ]
+
+        calls = {}
+
+        async def _fake_to_thread(fn, *args, **kwargs):
+            calls["args"] = args
+            calls["kwargs"] = kwargs
+            return fn(*args, **kwargs)
+
+        monkeypatch.setattr("gateway.run.asyncio.to_thread", _fake_to_thread)
+        monkeypatch.setattr(
+            "gateway.run.fetch_account_usage",
+            lambda provider, base_url=None, api_key=None: object(),
+        )
+        monkeypatch.setattr(
+            "gateway.run.render_account_usage_lines",
+            lambda snapshot, markdown=False: [
+                "📈 **Account limits**",
+                "Provider: openai-codex (Pro)",
+            ],
+        )
+
+        event = MagicMock()
+        result = await runner._handle_usage_command(event)
+
+        assert calls["args"] == ("openai-codex",)
+        assert calls["kwargs"]["base_url"] == "https://chatgpt.com/backend-api/codex"
+        assert "📊 **Session Info**" in result
+        assert "📈 **Account limits**" in result

tests/test_account_usage.py

@@ -0,0 +1,203 @@
+from datetime import datetime, timezone
+
+from agent.account_usage import (
+    AccountUsageSnapshot,
+    AccountUsageWindow,
+    fetch_account_usage,
+    render_account_usage_lines,
+)
+
+
+class _Response:
+    def __init__(self, payload, status_code=200):
+        self._payload = payload
+        self.status_code = status_code
+
+    def raise_for_status(self):
+        if self.status_code >= 400:
+            raise RuntimeError(f"HTTP {self.status_code}")
+
+    def json(self):
+        return self._payload
+
+
+class _Client:
+    def __init__(self, payload):
+        self._payload = payload
+
+    def __enter__(self):
+        return self
+
+    def __exit__(self, exc_type, exc, tb):
+        return False
+
+    def get(self, url, headers=None):
+        return _Response(self._payload)
+
+
+class _RoutingClient:
+    def __init__(self, payloads):
+        self._payloads = payloads
+
+    def __enter__(self):
+        return self
+
+    def __exit__(self, exc_type, exc, tb):
+        return False
+
+    def get(self, url, headers=None):
+        return _Response(self._payloads[url])
+
+
+def test_fetch_account_usage_codex(monkeypatch):
+    monkeypatch.setattr(
+        "agent.account_usage.resolve_codex_runtime_credentials",
+        lambda refresh_if_expiring=True: {
+            "provider": "openai-codex",
+            "base_url": "https://chatgpt.com/backend-api/codex",
+            "api_key": "***",
+        },
+    )
+    monkeypatch.setattr(
+        "agent.account_usage._read_codex_tokens",
+        lambda: {"tokens": {"account_id": "acct_123"}},
+    )
+    monkeypatch.setattr(
+        "agent.account_usage.httpx.Client",
+        lambda timeout=15.0: _Client(
+            {
+                "plan_type": "pro",
+                "rate_limit": {
+                    "primary_window": {
+                        "used_percent": 15,
+                        "reset_at": 1_900_000_000,
+                        "limit_window_seconds": 18000,
+                    },
+                    "secondary_window": {
+                        "used_percent": 40,
+                        "reset_at": 1_900_500_000,
+                        "limit_window_seconds": 604800,
+                    },
+                },
+                "credits": {"has_credits": True, "balance": 12.5},
+            }
+        ),
+    )
+
+    snapshot = fetch_account_usage("openai-codex")
+
+    assert snapshot is not None
+    assert snapshot.plan == "Pro"
+    assert len(snapshot.windows) == 2
+    assert snapshot.windows[0].label == "Session"
+    assert snapshot.windows[0].used_percent == 15.0
+    assert snapshot.windows[0].reset_at == datetime.fromtimestamp(1_900_000_000, tz=timezone.utc)
+    assert "Credits balance: $12.50" in snapshot.details
+
+
+def test_render_account_usage_lines_includes_reset_and_provider():
+    snapshot = AccountUsageSnapshot(
+        provider="openai-codex",
+        source="usage_api",
+        fetched_at=datetime.now(timezone.utc),
+        plan="Pro",
+        windows=(
+            AccountUsageWindow(
+                label="Session",
+                used_percent=25,
+                reset_at=datetime.now(timezone.utc),
+            ),
+        ),
+        details=("Credits balance: $9.99",),
+    )
+    lines = render_account_usage_lines(snapshot)
+
+    assert lines[0] == "📈 Account limits"
+    assert "openai-codex (Pro)" in lines[1]
+    assert "Session: 75% remaining (25% used)" in lines[2]
+    assert "Credits balance: $9.99" in lines[3]
+
+
+def test_fetch_account_usage_openrouter_uses_limit_remaining_and_ignores_deprecated_rate_limit(monkeypatch):
+    monkeypatch.setattr(
+        "agent.account_usage.resolve_runtime_provider",
+        lambda requested, explicit_base_url=None, explicit_api_key=None: {
+            "provider": "openrouter",
+            "base_url": "https://openrouter.ai/api/v1",
+            "api_key": "***",
+        },
+    )
+    monkeypatch.setattr(
+        "agent.account_usage.httpx.Client",
+        lambda timeout=10.0: _RoutingClient(
+            {
+                "https://openrouter.ai/api/v1/credits": {
+                    "data": {"total_credits": 300.0, "total_usage": 10.92}
+                },
+                "https://openrouter.ai/api/v1/key": {
+                    "data": {
+                        "limit": 100.0,
+                        "limit_remaining": 70.0,
+                        "limit_reset": "monthly",
+                        "usage": 12.5,
+                        "usage_daily": 0.5,
+                        "usage_weekly": 2.0,
+                        "usage_monthly": 8.0,
+                        "rate_limit": {"requests": -1, "interval": "10s"},
+                    }
+                },
+            }
+        ),
+    )
+
+    snapshot = fetch_account_usage("openrouter")
+
+    assert snapshot is not None
+    assert snapshot.windows == (
+        AccountUsageWindow(
+            label="API key quota",
+            used_percent=30.0,
+            detail="$70.00 of $100.00 remaining • resets monthly",
+        ),
+    )
+    assert "Credits balance: $289.08" in snapshot.details
+    assert "API key usage: $12.50 total • $0.50 today • $2.00 this week • $8.00 this month" in snapshot.details
+    assert all("-1 requests / 10s" not in line for line in render_account_usage_lines(snapshot))
+
+
+def test_fetch_account_usage_openrouter_omits_quota_window_when_key_has_no_limit(monkeypatch):
+    monkeypatch.setattr(
+        "agent.account_usage.resolve_runtime_provider",
+        lambda requested, explicit_base_url=None, explicit_api_key=None: {
+            "provider": "openrouter",
+            "base_url": "https://openrouter.ai/api/v1",
+            "api_key": "***",
+        },
+    )
+    monkeypatch.setattr(
+        "agent.account_usage.httpx.Client",
+        lambda timeout=10.0: _RoutingClient(
+            {
+                "https://openrouter.ai/api/v1/credits": {
+                    "data": {"total_credits": 100.0, "total_usage": 25.5}
+                },
+                "https://openrouter.ai/api/v1/key": {
+                    "data": {
+                        "limit": None,
+                        "limit_remaining": None,
+                        "usage": 25.5,
+                        "usage_daily": 1.25,
+                        "usage_weekly": 4.5,
+                        "usage_monthly": 18.0,
+                    }
+                },
+            }
+        ),
+    )
+
+    snapshot = fetch_account_usage("openrouter")
+
+    assert snapshot is not None
+    assert snapshot.windows == ()
+    assert "Credits balance: $74.50" in snapshot.details
+    assert "API key usage: $25.50 total • $1.25 today • $4.50 this week • $18.00 this month" in snapshot.details

tests/tools/test_voice_cli_integration.py

@@ -4,31 +4,13 @@ state management, streaming TTS activation, voice message prefix, _vprint."""
 import ast
 import os
 import queue
-import sys
 import threading
-import types
 from types import SimpleNamespace
 from unittest.mock import MagicMock, patch
 
 import pytest
 
 
-def _ensure_cli_import_shims():
-    sys.modules.setdefault(
-        "agent.auxiliary_client",
-        types.SimpleNamespace(
-            call_llm=lambda *args, **kwargs: "",
-            async_call_llm=lambda *args, **kwargs: "",
-            extract_content_or_reasoning=lambda *args, **kwargs: "",
-            resolve_provider_client=lambda *args, **kwargs: (None, None, None, None),
-            get_async_text_auxiliary_client=lambda *args, **kwargs: None,
-        ),
-    )
-
-
-_ensure_cli_import_shims()
-
-
 def _make_voice_cli(**overrides):
     """Create a minimal HermesCLI with only voice-related attrs initialized.
 
@@ -36,7 +18,6 @@ def _make_voice_cli(**overrides):
     needed.  Only the voice state attributes (from __init__ lines 3749-3758)
     are populated.
     """
-    _ensure_cli_import_shims()
     from cli import HermesCLI
 
     cli = HermesCLI.__new__(HermesCLI)
@@ -952,58 +933,6 @@ class TestEnableVoiceModeReal:
         assert cli._voice_mode is True
 
 
-class TestVoiceBeepConfigReal:
-    """Tests the CLI voice beep toggle."""
-
-    @patch("hermes_cli.config.load_config", return_value={"voice": {}})
-    def test_beeps_enabled_by_default(self, _cfg):
-        cli = _make_voice_cli()
-        assert cli._voice_beeps_enabled() is True
-
-    @patch("hermes_cli.config.load_config", return_value={"voice": {"beep_enabled": False}})
-    def test_beeps_can_be_disabled(self, _cfg):
-        cli = _make_voice_cli()
-        assert cli._voice_beeps_enabled() is False
-
-    @patch("cli._cprint")
-    @patch("cli.threading.Thread")
-    @patch("tools.voice_mode.play_beep")
-    @patch("tools.voice_mode.create_audio_recorder")
-    @patch(
-        "tools.voice_mode.check_voice_requirements",
-        return_value={
-            "available": True,
-            "audio_available": True,
-            "stt_available": True,
-            "details": "OK",
-            "missing_packages": [],
-        },
-    )
-    @patch(
-        "hermes_cli.config.load_config",
-        return_value={
-            "voice": {
-                "beep_enabled": False,
-                "silence_threshold": 200,
-                "silence_duration": 3.0,
-            }
-        },
-    )
-    def test_start_recording_skips_beep_when_disabled(
-        self, _cfg, _req, mock_create, mock_beep, mock_thread, _cp
-    ):
-        recorder = MagicMock()
-        recorder.supports_silence_autostop = True
-        mock_create.return_value = recorder
-        mock_thread.return_value = MagicMock(start=MagicMock())
-
-        cli = _make_voice_cli()
-        cli._voice_start_recording()
-
-        recorder.start.assert_called_once()
-        mock_beep.assert_not_called()
-
-
 class TestDisableVoiceModeReal:
     """Tests _disable_voice_mode with real CLI instance."""
 
@@ -1158,16 +1087,6 @@ class TestVoiceStopAndTranscribeReal:
         cli._voice_stop_and_transcribe()
         assert cli._pending_input.empty()
 
-    @patch("cli._cprint")
-    @patch("hermes_cli.config.load_config", return_value={"voice": {"beep_enabled": False}})
-    @patch("tools.voice_mode.play_beep")
-    def test_no_speech_detected_skips_beep_when_disabled(self, mock_beep, _cfg, _cp):
-        recorder = MagicMock()
-        recorder.stop.return_value = None
-        cli = _make_voice_cli(_voice_recording=True, _voice_recorder=recorder)
-        cli._voice_stop_and_transcribe()
-        mock_beep.assert_not_called()
-
     @patch("cli._cprint")
     @patch("cli.os.unlink")
     @patch("cli.os.path.isfile", return_value=True)
@@ -1237,18 +1156,12 @@ class TestVoiceStopAndTranscribeReal:
     @patch("cli._cprint")
     @patch("tools.voice_mode.play_beep")
     def test_continuous_restarts_on_no_speech(self, _beep, _cp):
-        import time
-
         recorder = MagicMock()
         recorder.stop.return_value = None
         cli = _make_voice_cli(_voice_recording=True, _voice_recorder=recorder,
                               _voice_continuous=True)
         cli._voice_start_recording = MagicMock()
         cli._voice_stop_and_transcribe()
-        for _ in range(50):
-            if cli._voice_start_recording.call_count:
-                break
-            time.sleep(0.01)
         cli._voice_start_recording.assert_called_once()
 
     @patch("cli._cprint")

website/docs/user-guide/features/voice-mode.md

@@ -149,7 +149,7 @@ Two-stage algorithm detects when you've finished speaking:
 
 If no speech is detected at all for 15 seconds, recording stops automatically.
 
-Both `silence_threshold` and `silence_duration` are configurable in `config.yaml`. You can also disable the record start/stop beeps with `voice.beep_enabled: false`.
+Both `silence_threshold` and `silence_duration` are configurable in `config.yaml`.
 
 ### Streaming TTS
 
@@ -383,7 +383,6 @@ voice:
   record_key: "ctrl+b"            # Key to start/stop recording
   max_recording_seconds: 120       # Maximum recording length
   auto_tts: false                  # Auto-enable TTS when voice mode starts
-  beep_enabled: true               # Play record start/stop beeps
   silence_threshold: 200           # RMS level (0-32767) below which counts as silence
   silence_duration: 3.0            # Seconds of silence before auto-stop