diff --git a/CHANGELOG.md b/CHANGELOG.md
index 09db7ea1..56ab74be 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -7,6 +7,30 @@
- New WebUI sessions no longer persist `display.personality` into per-session `Session.personality`; only explicit personality changes remain durable, preventing stale global display defaults from overriding profile-scoped session behavior. Closes #2845.
+## [v0.51.129] — 2026-05-24 — Release DA (stage-batch11 — 4-PR feature + perf batch)
+
+### Performance
+
+- **PR #2836** by @v2psv — HTTP/1.1 keep-alive for WebUI responses. Bumps `Handler.protocol_version` from the HTTP/1.0 default to `HTTP/1.1` so browsers can reuse TCP connections across normal API and static-file requests. Adds explicit `Content-Length` headers to hand-written responses that weren't already using shared `j()` / `t()` helpers. Adds `Content-Length: 0` to empty redirect / range-error responses. Switches SSE-style streaming endpoints from `Connection: keep-alive` to `Connection: close` (keep-alive is only safe when the response body is framed; SSE bodies have no fixed length). Significant first-paint / session-open improvements on high-RTT / VPN / proxied paths — author reports ~47% faster first paint and ~30-40% improvements on panel-load flows on a typical remote-host setup.
+
+ **Opus pre-release advisor caught one missing framing site** in the on-the-fly folder ZIP download path (`/api/folder/download`): the body has no known length, doesn't use chunked encoding, and was relying on HTTP/1.0 connection-close-equals-EOF. Under HTTP/1.1 this would have left clients hanging waiting for the next response after the central-directory bytes finished. Patched inline before tag: add `Connection: close` header to mirror the SSE-endpoint pattern. Opus verified this was the ONLY remaining streaming response in the codebase that needed the header — all 12 hand-written response paths + 8 SSE streams + j()/t() helpers + auth flow were already correctly framed by the PR.
+
+### Added
+
+- **PR #2680** by @mccxj — Auxiliary Models settings card in Settings → Preferences. Lets users configure per-task model routing for 9 canonical side-task slots: vision, web extract, compression, session search, approval, MCP tool reasoning, title generation, skills hub, curator. Each slot exposes a provider dropdown + model dropdown plus an "auto (use main model)" / "auto (use provider default)" pair so users can keep aux routing implicit when they don't care. New endpoints: `GET /api/model/auxiliary` returns current assignments; `POST /api/model/set` writes assignments (`scope=auxiliary` for aux slots, `scope=main` for the default chat model) and supports `task="__reset__"` to reset all slots back to auto. 16 new i18n keys added across all 12 locales (en, it, ja, ru, es, de, zh, zh-Hant, pt, ko, fr, tr — Turkish translations added in-stage to cover the sibling-PR collision with v0.51.127's Turkish locale baseline). 24 source-level test assertions covering HTML structure, JS logic, i18n parity, and route registration.
+
+- **PR #2842** by @AJV20 — PWA polish for installed launches. New `static/pwa-startup.js` is loaded synchronously in `` before the main UI bundle, so the page knows whether it's running standalone / in-browser / on iOS / offline before first paint. Marks `pwa-standalone`, `pwa-browser`, `pwa-ios`, `pwa-offline`, and short-lived `pwa-resumed` classes on ``. Exposes `window.HermesPWA.{isStandalone, syncMode, launchAction, promptInstall}` helpers and captures `beforeinstallprompt` / `appinstalled` early enough that any future install-prompt UI can chain off them. Manifest gains app identity / scope / `display_override` (`window-controls-overlay` → `standalone` → `minimal-ui`) and a "New conversation" PWA App Shortcut. Service worker pre-caches the startup helper, switches navigation and shell-asset fetches to `cache: 'no-store'` before falling back to CacheStorage. Boot path wires `?source=pwa&action=new-chat` to start a fresh chat instead of reopening the last saved session. The viewport meta now sets `maximum-scale=1, user-scalable=no` for native-feel — acknowledged trade-off against WCAG 2.1 1.4.4 (Resize text), intentionally kept for the PWA-installed feel of this user base.
+
+- **PR #2794** by @Michaelyklam — Runtime adapter route selection harness. Routes explicit adapter-mode chat starts through `build_runtime_adapter(...)` and keeps `legacy-direct` as the default `/api/chat/start` path. Continues the #1925 RFC slice progression: this is slice 4e, the default-off chat-start route-selection seam. Returns a bounded `501 Not Configured` response when `runner-local` is explicitly selected before a supervised runner client exists, instead of silently starting a legacy WebUI-owned run. New `_chat_start_response_from_run_start(...)` helper whitelists legacy-compatible chat-start response fields and keeps adapter-internal `run_id`, `status`, and `active_controls` out of public responses. Updates `docs/rfcs/hermes-run-adapter-contract.md` to mark #2744 shipped and define slice 4e.
+
+### Notes
+
+- Full pytest: **6,467 passed / 6 skipped / 3 xpassed / 8 subtests passed**.
+- Opus pre-release advisor reviewed all 7 risk areas (HTTP framing surface completeness, PWA startup ordering, sibling-PR `api/routes.py` interaction, service worker cache invalidation, viewport-meta trade-off, runtime adapter response shape, locale-counter brittleness). Verdict: **1 MUST-FIX patched inline** (folder ZIP `Connection: close` header), **0 inline SHOULD-FIX**, 1 follow-up suggested (`set_auxiliary_model` could validate `task` against `AUX_TASK_SLOTS` whitelist — auth-gated, low severity, filing as follow-up).
+- Agent self-verified: protocol_version bumped, SSE Connection-close + Content-Length plumbing, Auxiliary Models API surface (config + endpoints + frontend), PWA helpers + manifest shortcuts + display_override, Runtime adapter wiring + whitelisting, i18n parity for all 12 locales on the 16 new aux keys.
+- Browser-verified at 1920×1080: Auxiliary Models card renders correctly under Settings → Preferences, 9 task slots with provider/model dropdowns, "Reset all to auto" button, layout consistent with surrounding Settings cards, no clutter or clipping. PWA classes populate on `` and HermesPWA namespace populates with 4 helpers as expected.
+- In-stage commits added Turkish translations for #2680's 16 `settings_aux_*` / `settings_label_auxiliary_models` / `settings_desc_auxiliary_models` keys to close the sibling-collision gap with v0.51.127's Turkish locale (#2772). Bumped `test_auxiliary_models_settings.py::test_all_locales_have_auxiliary_keys` from `count == 11` to `count == 12` (the locale set grew when Turkish landed).
+
## [v0.51.128] — 2026-05-24 — Release CZ (stage-batch10 — 2-PR perf + correctness batch)
### Fixed
diff --git a/api/auth.py b/api/auth.py
index df5f0e4c..5a49516f 100644
--- a/api/auth.py
+++ b/api/auth.py
@@ -435,10 +435,12 @@ def check_auth(handler, parsed) -> bool:
return True
# Not authorized
if parsed.path.startswith('/api/'):
+ body = b'{"error":"Authentication required"}'
handler.send_response(401)
handler.send_header('Content-Type', 'application/json')
+ handler.send_header('Content-Length', str(len(body)))
handler.end_headers()
- handler.wfile.write(b'{"error":"Authentication required"}')
+ handler.wfile.write(body)
else:
handler.send_response(302)
# Pass the original path as ?next= so login.js redirects back after auth.
@@ -468,6 +470,7 @@ def check_auth(handler, parsed) -> bool:
# `?`, `&`, `=`) gets percent-encoded.
_next = _urlparse.quote(_path_with_query, safe='/')
handler.send_header('Location', 'login?next=' + _next)
+ handler.send_header('Content-Length', '0')
handler.end_headers()
return False
diff --git a/api/config.py b/api/config.py
index 293dc953..66bcb570 100644
--- a/api/config.py
+++ b/api/config.py
@@ -2185,6 +2185,112 @@ def set_hermes_default_model(model_id: str) -> dict:
return {"ok": True, "model": persisted_model}
+# ── Auxiliary model configuration ──────────────────────────────────────────
+
+# Canonical auxiliary task slots. Keep in sync with hermes_cli/config.py
+# DEFAULT_CONFIG["auxiliary"] and hermes_cli/web_server.py _AUX_TASK_SLOTS.
+AUX_TASK_SLOTS: tuple[str, ...] = (
+ "vision",
+ "web_extract",
+ "compression",
+ "session_search",
+ "skills_hub",
+ "approval",
+ "mcp",
+ "title_generation",
+ "curator",
+)
+
+
+def get_auxiliary_models() -> dict:
+ """Return current auxiliary task assignments from config.yaml.
+
+ Shape:
+ {
+ "tasks": [
+ {"task": "vision", "provider": "auto", "model": "", "base_url": ""},
+ ...
+ ],
+ "main": {"provider": "openrouter", "model": "anthropic/claude-opus-4.7"},
+ }
+ """
+ reload_config()
+ model_cfg = cfg.get("model", {})
+ if not isinstance(model_cfg, dict):
+ model_cfg = {}
+ main_provider = str(model_cfg.get("provider") or "").strip()
+ main_model = str(model_cfg.get("default") or model_cfg.get("name") or "").strip()
+
+ aux_cfg = cfg.get("auxiliary", {})
+ if not isinstance(aux_cfg, dict):
+ aux_cfg = {}
+
+ tasks = []
+ for slot in AUX_TASK_SLOTS:
+ entry = aux_cfg.get(slot, {})
+ if not isinstance(entry, dict):
+ entry = {}
+ tasks.append({
+ "task": slot,
+ "provider": str(entry.get("provider") or "auto").strip(),
+ "model": str(entry.get("model") or "").strip(),
+ "base_url": str(entry.get("base_url") or "").strip(),
+ })
+
+ return {
+ "tasks": tasks,
+ "main": {"provider": main_provider, "model": main_model},
+ }
+
+
+def set_auxiliary_model(task: str, provider: str, model: str) -> dict:
+ """Persist an auxiliary model assignment in config.yaml.
+
+ Special case: task='__reset__' clears all auxiliary slots.
+ """
+ config_path = _get_config_path()
+ with _cfg_lock:
+ config_data = _load_yaml_config_file(config_path)
+
+ if task == "__reset__":
+ # Per-slot reset: set each slot to auto, preserving extra fields
+ # (timeout, extra_body, api_key, base_url, download_timeout, etc.)
+ aux_cfg = config_data.get("auxiliary", {})
+ if not isinstance(aux_cfg, dict):
+ aux_cfg = {}
+ for slot in AUX_TASK_SLOTS:
+ slot_cfg = aux_cfg.get(slot, {})
+ if not isinstance(slot_cfg, dict):
+ slot_cfg = {}
+ slot_cfg["provider"] = "auto"
+ slot_cfg["model"] = ""
+ aux_cfg[slot] = slot_cfg
+ config_data["auxiliary"] = aux_cfg
+ else:
+ aux_cfg = config_data.get("auxiliary", {})
+ if not isinstance(aux_cfg, dict):
+ aux_cfg = {}
+ slot_cfg = aux_cfg.get(task, {})
+ if not isinstance(slot_cfg, dict):
+ slot_cfg = {}
+ slot_cfg["provider"] = provider or "auto"
+ slot_cfg["model"] = model or ""
+ if provider and (provider.startswith("custom:") or provider == "custom"):
+ try:
+ _, _, resolved_base_url = resolve_model_provider(model)
+ if resolved_base_url:
+ slot_cfg["base_url"] = str(resolved_base_url).strip().rstrip("/")
+ except Exception:
+ pass
+ aux_cfg[task] = slot_cfg
+ config_data["auxiliary"] = aux_cfg
+
+ _save_yaml_config_file(config_path, config_data)
+
+ reload_config()
+ return {"ok": True, "task": task, "provider": provider, "model": model}
+
+
# ── TTL cache for get_available_models() ─────────────────────────────────────
_available_models_cache: dict | None = None
_available_models_cache_ts: float = 0.0
diff --git a/api/kanban_bridge.py b/api/kanban_bridge.py
index 63bef9cd..f0d5d261 100644
--- a/api/kanban_bridge.py
+++ b/api/kanban_bridge.py
@@ -1022,7 +1022,7 @@ def _handle_events_sse_stream(handler, parsed):
handler.send_header("Content-Type", "text/event-stream; charset=utf-8")
handler.send_header("Cache-Control", "no-cache")
handler.send_header("X-Accel-Buffering", "no")
- handler.send_header("Connection", "keep-alive")
+ handler.send_header("Connection", "close")
handler.end_headers()
# Send an initial frame so the client knows the connection is open
diff --git a/api/routes.py b/api/routes.py
index b6ee042c..54b8818a 100644
--- a/api/routes.py
+++ b/api/routes.py
@@ -3687,6 +3687,11 @@ def handle_get(handler, parsed) -> bool:
if parsed.path == "/api/models/live":
return _handle_live_models(handler, parsed)
+ # ── Auxiliary models (GET/POST) ──
+ if parsed.path == "/api/model/auxiliary":
+ from api.config import get_auxiliary_models
+ return j(handler, get_auxiliary_models())
+
if parsed.path == "/api/dashboard/status":
from api import dashboard_probe
@@ -4857,6 +4862,25 @@ def handle_post(handler, parsed) -> bool:
except RuntimeError as e:
return bad(handler, str(e), 500)
+ # ── Auxiliary model set (POST) ──
+ if parsed.path == "/api/model/set":
+ scope = str(body.get("scope") or "").strip()
+ task = str(body.get("task") or "").strip()
+ provider = str(body.get("provider") or "auto").strip()
+ model = str(body.get("model") or "").strip()
+ if scope == "auxiliary":
+ from api.config import set_auxiliary_model
+ try:
+ return j(handler, set_auxiliary_model(task, provider, model))
+ except Exception as exc:
+ return bad(handler, str(exc), status=400)
+ if scope == "main":
+ try:
+ return j(handler, set_hermes_default_model(model))
+ except ValueError as exc:
+ return bad(handler, str(exc), status=400)
+ return bad(handler, f"unknown scope: {scope}", status=400)
+
# ── Providers (POST) ──
if parsed.path == "/api/providers":
provider_id = (body.get("provider") or "").strip().lower()
@@ -6216,13 +6240,15 @@ def handle_post(handler, parsed) -> bool:
_record_login_attempt(client_ip)
return bad(handler, "Invalid password", 401)
cookie_val = create_session()
+ body = json.dumps({"ok": True}).encode()
handler.send_response(200)
handler.send_header("Content-Type", "application/json")
+ handler.send_header("Content-Length", str(len(body)))
handler.send_header("Cache-Control", "no-store")
_security_headers(handler)
set_auth_cookie(handler, cookie_val)
handler.end_headers()
- handler.wfile.write(json.dumps({"ok": True}).encode())
+ handler.wfile.write(body)
return True
if parsed.path == "/api/auth/logout":
@@ -6231,13 +6257,15 @@ def handle_post(handler, parsed) -> bool:
cookie_val = parse_cookie(handler)
if cookie_val:
invalidate_session(cookie_val)
+ body = json.dumps({"ok": True}).encode()
handler.send_response(200)
handler.send_header("Content-Type", "application/json")
+ handler.send_header("Content-Length", str(len(body)))
handler.send_header("Cache-Control", "no-store")
_security_headers(handler)
clear_auth_cookie(handler)
handler.end_headers()
- handler.wfile.write(json.dumps({"ok": True}).encode())
+ handler.wfile.write(body)
return True
# ── Checkpoints / Rollback (POST) ──
@@ -6577,7 +6605,7 @@ def _handle_sse_stream(handler, parsed):
handler.send_header("Content-Type", "text/event-stream; charset=utf-8")
handler.send_header("Cache-Control", "no-cache")
handler.send_header("X-Accel-Buffering", "no")
- handler.send_header("Connection", "keep-alive")
+ handler.send_header("Connection", "close")
handler.end_headers()
try:
_replay_run_journal(handler, stream_id, _parse_run_journal_after_seq(qs))
@@ -6589,7 +6617,7 @@ def _handle_sse_stream(handler, parsed):
handler.send_header("Content-Type", "text/event-stream; charset=utf-8")
handler.send_header("Cache-Control", "no-cache")
handler.send_header("X-Accel-Buffering", "no")
- handler.send_header("Connection", "keep-alive")
+ handler.send_header("Connection", "close")
handler.end_headers()
try:
while True:
@@ -6720,7 +6748,7 @@ def _handle_terminal_output(handler, parsed):
handler.send_header("Content-Type", "text/event-stream; charset=utf-8")
handler.send_header("Cache-Control", "no-cache")
handler.send_header("X-Accel-Buffering", "no")
- handler.send_header("Connection", "keep-alive")
+ handler.send_header("Connection", "close")
handler.end_headers()
try:
while True:
@@ -6798,7 +6826,7 @@ def _handle_gateway_sse_stream(handler, parsed):
handler.send_header('Content-Type', 'text/event-stream; charset=utf-8')
handler.send_header('Cache-Control', 'no-cache')
handler.send_header('X-Accel-Buffering', 'no')
- handler.send_header('Connection', 'keep-alive')
+ handler.send_header('Connection', 'close')
handler.end_headers()
q = watcher.subscribe()
@@ -6831,7 +6859,7 @@ def _handle_session_events_stream(handler):
handler.send_header('Content-Type', 'text/event-stream; charset=utf-8')
handler.send_header('Cache-Control', 'no-cache')
handler.send_header('X-Accel-Buffering', 'no')
- handler.send_header('Connection', 'keep-alive')
+ handler.send_header('Connection', 'close')
handler.end_headers()
q = subscribe_session_events()
@@ -6917,6 +6945,7 @@ def _serve_file_bytes(handler, target: Path, mime: str, disposition: str, cache_
handler.send_response(416)
handler.send_header("Content-Range", f"bytes */{file_size}")
handler.send_header("Accept-Ranges", "bytes")
+ handler.send_header("Content-Length", "0")
_security_headers(handler)
handler.end_headers()
return True
@@ -7027,10 +7056,12 @@ def _handle_media(handler, parsed):
if is_auth_enabled():
cv = parse_cookie(handler)
if not (cv and verify_session(cv)):
+ body = b'{"error":"Authentication required"}'
handler.send_response(401)
handler.send_header("Content-Type", "application/json")
+ handler.send_header("Content-Length", str(len(body)))
handler.end_headers()
- handler.wfile.write(b'{"error":"Authentication required"}')
+ handler.wfile.write(body)
return
qs = parse_qs(parsed.query)
@@ -7264,6 +7295,15 @@ def _handle_folder_download(handler, parsed):
_content_disposition_value("attachment", zip_name),
)
handler.send_header("Cache-Control", "no-store")
+ # Under HTTP/1.1 (Handler.protocol_version, see server.py post-#2836)
+ # a response with no Content-Length and no Transfer-Encoding requires
+ # Connection: close so the client knows the body ends at FIN. The ZIP
+ # is built on-the-fly so we cannot send Content-Length up front; mirror
+ # the SSE-endpoint pattern #2836 uses. Without this header the client
+ # hangs waiting for the next pipelined response after the central
+ # directory bytes finish. Caught by Opus pre-release advisor on
+ # stage-batch11.
+ handler.send_header("Connection", "close")
handler.end_headers()
written = 0
@@ -7389,7 +7429,7 @@ def _handle_approval_sse_stream(handler, parsed):
handler.send_header('Content-Type', 'text/event-stream; charset=utf-8')
handler.send_header('Cache-Control', 'no-cache')
handler.send_header('X-Accel-Buffering', 'no')
- handler.send_header('Connection', 'keep-alive')
+ handler.send_header('Connection', 'close')
handler.end_headers()
from api.streaming import _sse
@@ -7490,7 +7530,7 @@ def _handle_clarify_sse_stream(handler, parsed):
handler.send_header('Content-Type', 'text/event-stream; charset=utf-8')
handler.send_header('Cache-Control', 'no-cache')
handler.send_header('X-Accel-Buffering', 'no')
- handler.send_header('Connection', 'keep-alive')
+ handler.send_header('Connection', 'close')
handler.end_headers()
from api.streaming import _sse
@@ -8463,6 +8503,41 @@ def _start_chat_stream_for_session(
return response
+def _runtime_runner_client_factory():
+ """Return the runner-local client when a supervised backend exists.
+
+ Slice 4d wires the `/api/chat/start` selection point without silently falling
+ back to the legacy in-process runtime when `runner-local` is explicitly
+ requested. The supervised runner backend itself is intentionally not created
+ in this helper yet; a later slice can replace this factory with the concrete
+ client while keeping the route contract stable.
+ """
+ raise NotImplementedError("runner-local chat backend is not configured")
+
+
+def _chat_start_response_from_run_start(result):
+ """Expose only the legacy browser-facing chat-start response fields."""
+ payload = dict(getattr(result, "payload", {}) or {})
+ response = {}
+ for key in (
+ "stream_id",
+ "session_id",
+ "pending_started_at",
+ "turn_id",
+ "title",
+ "effective_model",
+ "effective_model_provider",
+ "error",
+ "active_stream_id",
+ "_status",
+ ):
+ if key in payload:
+ response[key] = payload[key]
+ response.setdefault("stream_id", result.stream_id)
+ response.setdefault("session_id", result.session_id)
+ return response
+
+
def _runtime_adapter_goal_action(goal_args: str) -> str:
"""Return the bounded RuntimeAdapter goal action for WebUI /goal args."""
action = str(goal_args or "").strip().lower()
@@ -8672,10 +8747,12 @@ def _handle_chat_start(handler, body, diag=None):
from api.runtime_adapter import (
LegacyJournalRuntimeAdapter,
StartRunRequest,
+ build_runtime_adapter,
runtime_adapter_enabled,
+ runtime_adapter_runner_enabled,
)
- if runtime_adapter_enabled():
+ if runtime_adapter_enabled() or runtime_adapter_runner_enabled():
def _legacy_start_run(request: StartRunRequest) -> dict:
return _start_chat_stream_for_session(
s,
@@ -8688,23 +8765,32 @@ def _handle_chat_start(handler, body, diag=None):
diag=diag,
)
- adapter = LegacyJournalRuntimeAdapter(start_run_delegate=_legacy_start_run)
- result = adapter.start_run(
- StartRunRequest(
- session_id=s.session_id,
- message=msg,
- attachments=attachments,
- workspace=workspace,
- profile=getattr(s, "profile", None),
- provider=model_provider,
- model=model,
- source="webui",
- metadata={"route": "/api/chat/start"},
+ def _legacy_adapter_factory():
+ return LegacyJournalRuntimeAdapter(start_run_delegate=_legacy_start_run)
+
+ try:
+ adapter = build_runtime_adapter(
+ legacy_adapter_factory=_legacy_adapter_factory,
+ runner_client_factory=_runtime_runner_client_factory,
)
- )
- response = dict(result.payload)
- response.setdefault("stream_id", result.stream_id)
- response.setdefault("session_id", result.session_id)
+ if adapter is None:
+ raise NotImplementedError("runtime adapter selection returned no adapter")
+ result = adapter.start_run(
+ StartRunRequest(
+ session_id=s.session_id,
+ message=msg,
+ attachments=attachments,
+ workspace=workspace,
+ profile=getattr(s, "profile", None),
+ provider=model_provider,
+ model=model,
+ source="webui",
+ metadata={"route": "/api/chat/start"},
+ )
+ )
+ except NotImplementedError as exc:
+ return j(handler, {"error": str(exc)}, status=501)
+ response = _chat_start_response_from_run_start(result)
else:
response = _start_chat_stream_for_session(
s,
diff --git a/docs/rfcs/hermes-run-adapter-contract.md b/docs/rfcs/hermes-run-adapter-contract.md
index 9732044e..71ef4da2 100644
--- a/docs/rfcs/hermes-run-adapter-contract.md
+++ b/docs/rfcs/hermes-run-adapter-contract.md
@@ -4,7 +4,7 @@
- **Author:** @Michaelyklam
- **Updated by:** @franksong2702
- **Created:** 2026-05-11
-- **Revised:** 2026-05-21
+- **Revised:** 2026-05-23
- **Tracking issue:** [#1925](https://github.com/nesquena/hermes-webui/issues/1925)
## Credit and Scope
@@ -52,7 +52,7 @@ The immediate goal is not to build a sidecar. The immediate goal is to define th
browser contract, classify current runtime state, and gate the first reversible
journal slice.
-## Current Gate State — 2026-05-21
+## Current Gate State — 2026-05-23
Slice 1 is now past the first active validation gate:
@@ -104,11 +104,14 @@ adapter-seam work:
`runner-local` adapter selection point and `build_runtime_adapter(...)`
factory wiring around an injected runner client. Live browser chat routes still
stay on the legacy backend, and no supervised runner process exists yet.
-- The next implementation gate is a supervised/local runner backend proposal and
- route-selection harness. It must stay default-off, keep legacy fallback intact,
- pass explicit profile/workspace/model payloads instead of mutating WebUI
- process globals, and avoid recreating `STREAMS` / `CANCEL_FLAGS` / approval
- queues / clarify queues under new names.
+- #2744 shipped the Slice 4d supervised runner route gate in v0.51.108.
+- The next implementation slice is a default-off runner route-selection harness
+ for `/api/chat/start`. It should only engage when `runner-local` is explicitly
+ selected, return a bounded not-configured error until a supervised runner
+ client exists, keep `legacy-direct` / `legacy-journal` fallback intact, pass
+ explicit profile/workspace/model payloads instead of mutating WebUI process
+ globals, and avoid recreating `STREAMS` / `CANCEL_FLAGS` / approval queues /
+ clarify queues under new names.
The next gate is runner-backend plumbing, not queue implementation
by default. Queue / continue routing should only move before Slice 4 if a future
@@ -843,6 +846,10 @@ Non-goals for Slice 4c:
#### Slice 4d: Supervised runner backend route gate
+Status as of 2026-05-23: shipped in v0.51.108 via #2744. The gate remains a
+docs/test contract: it defines the default-off route-selection requirements but
+does not itself route live chat to a runner backend.
+
After `runner-local` selection exists, the next reviewable gate should define the
first supervised/local runner backend and the route-selection harness before live
browser chat can use it. This is still a contract/test slice first: no default-on
@@ -896,6 +903,51 @@ Non-goals for Slice 4d:
- no broad UI/product surface migration; WebUI remains the rich workbench while
only execution ownership moves.
+#### Slice 4e: Default-off runner chat-start route-selection harness
+
+The first implementation after the Slice 4d gate should wire the
+`/api/chat/start` selection point to the existing `RuntimeAdapter` factory
+without adding a supervised runner process yet. The harness must make the
+selection behavior explicit: `legacy-direct` stays default, `legacy-journal`
+continues to delegate to the legacy in-process stream path, and `runner-local`
+does not silently fall back to legacy when no runner client is configured.
+
+Scope:
+
+- route `/api/chat/start` through `build_runtime_adapter(...)` when an adapter
+ mode is explicitly selected;
+- keep the successful browser response whitelisted to legacy-compatible fields
+ such as `stream_id`, `session_id`, `pending_started_at`, `turn_id`, `title`,
+ and effective model/provider metadata;
+- return a bounded not-configured error for `runner-local` until a supervised
+ runner client/backend lands;
+- pass the existing explicit `StartRunRequest` payload fields across the seam.
+
+Acceptance tests for Slice 4e:
+
+1. **Default remains legacy-direct.** With no adapter env var, `/api/chat/start`
+ keeps using `_start_chat_stream_for_session(...)` directly.
+2. **Legacy-journal remains behavior-preserving.** The flagged legacy adapter
+ still delegates to the same stream-start helper and preserves the public
+ response shape.
+3. **Runner-local does not fallback silently.** If `runner-local` is selected but
+ no runner client exists, the route returns a bounded error instead of starting
+ a WebUI-owned legacy run behind the operator's back.
+4. **No adapter-internal response drift.** `run_id`, `status`, and
+ `active_controls` remain internal until a later contract explicitly exposes
+ them.
+5. **No runtime-surrogate globals.** The harness does not add runner-owned stream,
+ cancel, approval, clarify, cached-agent, goal, or queue maps to the main WebUI
+ process.
+
+Non-goals for Slice 4e:
+
+- no supervised runner process yet;
+- no default-on runner mode;
+- no execution-survives-WebUI-restart claim for production chat turns;
+- no removal of `legacy-direct` or `legacy-journal`;
+- no server-side queue endpoint or queue scheduler just for adapter symmetry.
+
## First Meaningful Success Criteria
The first meaningful milestones are deliberately split.
diff --git a/server.py b/server.py
index e6636209..e6a2c65a 100644
--- a/server.py
+++ b/server.py
@@ -170,6 +170,13 @@ class QuietHTTPServer(ThreadingHTTPServer):
class Handler(BaseHTTPRequestHandler):
+ # HTTP/1.1 enables keep-alive connection reuse — major latency win on
+ # high-RTT links where every saved TCP handshake is 2×RTT. Each response
+ # MUST declare framing (Content-Length, Transfer-Encoding: chunked, or
+ # Connection: close) so the client knows where the message ends. Helpers
+ # j()/t() emit Content-Length; SSE/streaming endpoints emit
+ # Connection: close because the body has no terminator. See PR notes.
+ protocol_version = "HTTP/1.1"
timeout = 30 # seconds — kills idle/incomplete connections to prevent thread exhaustion
def setup(self):
diff --git a/static/boot.js b/static/boot.js
index 2b6d8347..b1c67dd6 100644
--- a/static/boot.js
+++ b/static/boot.js
@@ -1662,6 +1662,17 @@ function applyBotName(){
if(typeof fetchReasoningChip==='function') fetchReasoningChip();
if(typeof refreshProviderQuotaIndicator==='function') refreshProviderQuotaIndicator();
const urlSession=(typeof _sessionIdFromLocation==='function')?_sessionIdFromLocation():null;
+ const pwaLaunchAction=(window.HermesPWA&&typeof window.HermesPWA.launchAction==='function')
+ ? window.HermesPWA.launchAction()
+ : null;
+ if(pwaLaunchAction==='new-chat'){
+ try{
+ await newSession(true);
+ if(S.session) await _startBootModelDropdown();
+ S._bootReady=true;
+ syncTopbar();syncWorkspacePanelState();await renderSessionList();if(typeof startGatewaySSE==='function')startGatewaySSE();return;
+ }catch(e){console.warn('[pwa] new-chat launch action failed', e);}
+ }
const savedLocal=localStorage.getItem('hermes-webui-session');
const saved=urlSession||savedLocal;
if(saved){
diff --git a/static/i18n.js b/static/i18n.js
index c1ba196f..d103d996 100644
--- a/static/i18n.js
+++ b/static/i18n.js
@@ -548,6 +548,22 @@ const LOCALES = {
settings_save_btn: 'Save Settings',
settings_label_model: 'Default Model',
settings_desc_model: 'Used for new conversations. Existing conversations keep their selected model.',
+ settings_label_auxiliary_models: 'Auxiliary Models',
+ settings_desc_auxiliary_models: 'Side-task routing for vision, compression, title generation, etc. "Auto" uses your main chat model.',
+ settings_btn_reset_aux_models: 'Reset all to auto',
+ settings_btn_apply_aux_models: 'Apply changes',
+ settings_aux_provider_auto: 'use main model',
+ settings_aux_model_auto: 'auto (use provider default)',
+ settings_aux_model_custom: 'Custom model…',
+ settings_aux_model_custom_prompt: 'Enter model ID:',
+ settings_aux_loading: 'Loading auxiliary models…',
+ settings_aux_load_failed: 'Could not load auxiliary model settings. Make sure the agent API is available.',
+ settings_aux_reset_confirm_title: 'Reset auxiliary models?',
+ settings_aux_reset_confirm_msg: 'This will set all auxiliary tasks to auto (use main model).',
+ settings_aux_reset_done: 'Auxiliary models reset to auto',
+ settings_aux_save_failed: 'Failed to save auxiliary model',
+ settings_aux_saved: 'Auxiliary models updated',
+ settings_aux_no_changes: 'No changes to apply',
settings_label_send_key: 'Send Key',
settings_label_theme: 'Theme',
settings_label_skin: 'Skin',
@@ -1791,6 +1807,22 @@ const LOCALES = {
settings_save_btn: 'Salva Impostazioni',
settings_label_model: 'Modello Predefinito',
settings_desc_model: 'Usato per le nuove conversazioni. Le conversazioni esistenti mantengono il modello selezionato.',
+ settings_label_auxiliary_models: 'Modelli Ausiliari',
+ settings_desc_auxiliary_models: 'Routing per attività secondarie come visione, compressione, generazione titoli, ecc. "Auto" utilizza il modello chat principale.',
+ settings_btn_reset_aux_models: 'Ripristina tutto ad auto',
+ settings_btn_apply_aux_models: 'Applica modifiche',
+ settings_aux_provider_auto: 'usa modello principale',
+ settings_aux_model_auto: 'auto (usa predefinito del provider)',
+ settings_aux_model_custom: 'Modello personalizzato…',
+ settings_aux_model_custom_prompt: 'Inserisci ID modello:',
+ settings_aux_loading: 'Caricamento modelli ausiliari…',
+ settings_aux_load_failed: 'Impossibile caricare le impostazioni dei modelli ausiliari. Assicurati che l’API dell’agent sia disponibile.',
+ settings_aux_reset_confirm_title: 'Ripristinare modelli ausiliari?',
+ settings_aux_reset_confirm_msg: 'Questo imposterà tutte le attività ausiliarie ad auto (usa modello principale).',
+ settings_aux_reset_done: 'Modelli ausiliari ripristinati ad auto',
+ settings_aux_save_failed: 'Salvataggio del modello ausiliario non riuscito',
+ settings_aux_saved: 'Modelli ausiliari aggiornati',
+ settings_aux_no_changes: 'Nessuna modifica da applicare',
settings_label_send_key: 'Tasto Invio',
settings_label_theme: 'Tema',
settings_label_skin: 'Skin',
@@ -3026,6 +3058,22 @@ const LOCALES = {
settings_save_btn: '設定を保存',
settings_label_model: 'デフォルトモデル',
settings_desc_model: '新しい会話で使用されます。既存の会話は選択済みモデルを保持します。',
+ settings_label_auxiliary_models: '補助モデル',
+ settings_desc_auxiliary_models: 'ビジョン、圧縮、タイトル生成などの補助タスクのルーティング。「自動」はメインチャットモデルを使用します。',
+ settings_btn_reset_aux_models: 'すべて自動にリセット',
+ settings_btn_apply_aux_models: '変更を適用',
+ settings_aux_provider_auto: 'メインモデルを使用',
+ settings_aux_model_auto: '自動（プロバイダーのデフォルトを使用）',
+ settings_aux_model_custom: 'カスタムモデル…',
+ settings_aux_model_custom_prompt: 'モデルIDを入力：',
+ settings_aux_loading: '補助モデルを読み込み中…',
+ settings_aux_load_failed: '補助モデル設定を読み込めませんでした。エージェントAPIが利用可能であることを確認してください。',
+ settings_aux_reset_confirm_title: '補助モデルをリセットしますか？',
+ settings_aux_reset_confirm_msg: 'すべての補助タスクを自動（メインモデルを使用）に設定します。',
+ settings_aux_reset_done: '補助モデルを自動にリセットしました',
+ settings_aux_save_failed: '補助モデルの保存に失敗しました',
+ settings_aux_saved: '補助モデルを更新しました',
+ settings_aux_no_changes: '適用する変更はありません',
settings_label_send_key: '送信キー',
settings_label_theme: 'テーマ',
settings_label_skin: 'スキン',
@@ -4077,6 +4125,22 @@ const LOCALES = {
settings_save_btn: 'Сохранить настройки',
settings_label_model: 'Модель по умолчанию',
settings_desc_model: 'Используется для новых бесед. Существующие беседы сохраняют выбранную модель.',
+ settings_label_auxiliary_models: 'Вспомогательные модели',
+ settings_desc_auxiliary_models: 'Маршрутизация побочных задач: зрение, сжатие, генерация заголовков и т.д. «Авто» использует основную модель чата.',
+ settings_btn_reset_aux_models: 'Сбросить всё на авто',
+ settings_btn_apply_aux_models: 'Применить изменения',
+ settings_aux_provider_auto: 'использовать основную модель',
+ settings_aux_model_auto: 'авто (по умолчанию провайдера)',
+ settings_aux_model_custom: 'Пользовательская модель…',
+ settings_aux_model_custom_prompt: 'Введите ID модели:',
+ settings_aux_loading: 'Загрузка вспомогательных моделей…',
+ settings_aux_load_failed: 'Не удалось загрузить настройки вспомогательных моделей. Убедитесь, что API агента доступен.',
+ settings_aux_reset_confirm_title: 'Сбросить вспомогательные модели?',
+ settings_aux_reset_confirm_msg: 'Все вспомогательные задачи будут установлены на «авто» (использовать основную модель).',
+ settings_aux_reset_done: 'Вспомогательные модели сброшены на авто',
+ settings_aux_save_failed: 'Не удалось сохранить вспомогательную модель',
+ settings_aux_saved: 'Вспомогательные модели обновлены',
+ settings_aux_no_changes: 'Нет изменений для применения',
settings_label_send_key: 'Клавиша отправки',
settings_label_theme: 'Тема',
settings_label_language: 'Язык',
@@ -5231,6 +5295,22 @@ const LOCALES = {
settings_save_btn: 'Guardar configuración',
settings_label_model: 'Modelo predeterminado',
settings_desc_model: 'Se usa para conversaciones nuevas. Las conversaciones existentes conservan su modelo seleccionado.',
+ settings_label_auxiliary_models: 'Modelos auxiliares',
+ settings_desc_auxiliary_models: 'Enrutamiento para tareas secundarias como visión, compresión, generación de títulos, etc. «Auto» usa el modelo de chat principal.',
+ settings_btn_reset_aux_models: 'Restablecer todo a auto',
+ settings_btn_apply_aux_models: 'Aplicar cambios',
+ settings_aux_provider_auto: 'usar modelo principal',
+ settings_aux_model_auto: 'auto (usar predeterminado del proveedor)',
+ settings_aux_model_custom: 'Modelo personalizado…',
+ settings_aux_model_custom_prompt: 'Ingrese ID del modelo:',
+ settings_aux_loading: 'Cargando modelos auxiliares…',
+ settings_aux_load_failed: 'No se pudieron cargar las configuraciones de modelos auxiliares. Asegúrese de que la API del agente esté disponible.',
+ settings_aux_reset_confirm_title: '¿Restablecer modelos auxiliares?',
+ settings_aux_reset_confirm_msg: 'Esto establecerá todas las tareas auxiliares en auto (usar modelo principal).',
+ settings_aux_reset_done: 'Modelos auxiliares restablecidos a auto',
+ settings_aux_save_failed: 'Error al guardar el modelo auxiliar',
+ settings_aux_saved: 'Modelos auxiliares actualizados',
+ settings_aux_no_changes: 'Sin cambios para aplicar',
settings_label_send_key: 'Tecla de envío',
settings_label_theme: 'Tema',
settings_label_skin: 'Piel',
@@ -6388,6 +6468,22 @@ const LOCALES = {
settings_save_btn: 'Einstellungen speichern',
settings_label_model: 'Standard-Modell',
settings_desc_model: 'Wird für neue Chats verwendet. Bestehende Chats behalten ihr ausgewähltes Modell.',
+ settings_label_auxiliary_models: 'Hilfsmodelle',
+ settings_desc_auxiliary_models: 'Routing für Nebenaufgaben wie Vision, Komprimierung, Titelgenerierung usw. „Auto" verwendet das Haupt-Chat-Modell.',
+ settings_btn_reset_aux_models: 'Alle auf auto zurücksetzen',
+ settings_btn_apply_aux_models: 'Änderungen anwenden',
+ settings_aux_provider_auto: 'Hauptmodell verwenden',
+ settings_aux_model_auto: 'auto (Provider-Standard verwenden)',
+ settings_aux_model_custom: 'Benutzerdefiniertes Modell…',
+ settings_aux_model_custom_prompt: 'Modell-ID eingeben:',
+ settings_aux_loading: 'Hilfsmodelle werden geladen…',
+ settings_aux_load_failed: 'Hilfsmodelle-Einstellungen konnten nicht geladen werden. Stellen Sie sicher, dass die Agent-API verfügbar ist.',
+ settings_aux_reset_confirm_title: 'Hilfsmodelle zurücksetzen?',
+ settings_aux_reset_confirm_msg: 'Dies setzt alle Hilfsaufgaben auf auto (Hauptmodell verwenden).',
+ settings_aux_reset_done: 'Hilfsmodelle auf auto zurückgesetzt',
+ settings_aux_save_failed: 'Hilfsmodell konnte nicht gespeichert werden',
+ settings_aux_saved: 'Hilfsmodelle aktualisiert',
+ settings_aux_no_changes: 'Keine Änderungen anzuwenden',
settings_label_send_key: 'Sende-Taste',
settings_label_theme: 'Theme',
settings_label_skin: 'Skin',
@@ -7597,6 +7693,22 @@ const LOCALES = {
settings_save_btn: '保存设置',
settings_label_model: '默认模型',
settings_desc_model: '用于新对话。现有对话保持各自选定的模型。',
+ settings_label_auxiliary_models: '辅助模型',
+ settings_desc_auxiliary_models: '视觉分析、上下文压缩、标题生成等辅助任务的路由。"自动"表示使用主聊天模型。',
+ settings_btn_reset_aux_models: '全部重置为自动',
+ settings_btn_apply_aux_models: '应用更改',
+ settings_aux_provider_auto: '使用主模型',
+ settings_aux_model_auto: '自动（使用提供商默认）',
+ settings_aux_model_custom: '自定义模型…',
+ settings_aux_model_custom_prompt: '输入模型 ID：',
+ settings_aux_loading: '正在加载辅助模型…',
+ settings_aux_load_failed: '无法加载辅助模型设置，请确保 Agent API 可用。',
+ settings_aux_reset_confirm_title: '重置辅助模型？',
+ settings_aux_reset_confirm_msg: '这将把所有辅助任务设置为自动（使用主模型）。',
+ settings_aux_reset_done: '辅助模型已重置为自动',
+ settings_aux_save_failed: '辅助模型保存失败',
+ settings_aux_saved: '辅助模型已更新',
+ settings_aux_no_changes: '没有需要应用的更改',
settings_label_send_key: '发送快捷键',
settings_label_theme: '主题',
settings_label_skin: '皮肤',
@@ -8783,6 +8895,22 @@ const LOCALES = {
settings_save_btn: '\u5132\u5b58\u8a2d\u5b9a',
settings_label_model: '\u9ed8\u8a8d\u6a21\u578b',
settings_desc_model: '\u7528\u65bc\u65b0\u6703\u8a71\u3002\u73fe\u6709\u6703\u8a71\u6703\u4fdd\u7559\u5404\u81ea\u9078\u5b9a\u7684\u6a21\u578b\u3002',
+ settings_label_auxiliary_models: '\u8f14\u52a9\u6a21\u578b',
+ settings_desc_auxiliary_models: '\u8996\u89ba\u5206\u6790\u3001\u4e0a\u4e0b\u6587\u58d3\u7e2e\u3001\u6a19\u984c\u7522\u751f\u7b49\u8f14\u52a9\u4efb\u52d9\u7684\u8def\u7531\u3002\u300c\u81ea\u52d5\u300d\u8868\u793a\u4f7f\u7528\u4e3b\u804a\u5929\u6a21\u578b\u3002',
+ settings_btn_reset_aux_models: '\u5168\u90e8\u91cd\u7f6e\u70ba\u81ea\u52d5',
+ settings_btn_apply_aux_models: '\u61c9\u7528\u8b8a\u66f4',
+ settings_aux_provider_auto: '\u4f7f\u7528\u4e3b\u6a21\u578b',
+ settings_aux_model_auto: '\u81ea\u52d5\uff08\u4f7f\u7528\u63d0\u4f9b\u5546\u9810\u8a2d\uff09',
+ settings_aux_model_custom: '\u81ea\u8a02\u6a21\u578b\u2026',
+ settings_aux_model_custom_prompt: '\u8f38\u5165\u6a21\u578b ID\uff1a',
+ settings_aux_loading: '\u6b63\u5728\u8f09\u5165\u8f14\u52a9\u6a21\u578b\u2026',
+ settings_aux_load_failed: '\u7121\u6cd5\u8f09\u5165\u8f14\u52a9\u6a21\u578b\u8a2d\u5b9a\uff0c\u8acb\u78ba\u4fdd Agent API \u53ef\u7528\u3002',
+ settings_aux_reset_confirm_title: '\u91cd\u7f6e\u8f14\u52a9\u6a21\u578b\uff1f',
+ settings_aux_reset_confirm_msg: '\u9019\u5c07\u628a\u6240\u6709\u8f14\u52a9\u4efb\u52d9\u8a2d\u5b9a\u70ba\u81ea\u52d5\uff08\u4f7f\u7528\u4e3b\u6a21\u578b\uff09\u3002',
+ settings_aux_reset_done: '\u8f14\u52a9\u6a21\u578b\u5df2\u91cd\u7f6e\u70ba\u81ea\u52d5',
+ settings_aux_save_failed: '\u8f14\u52a9\u6a21\u578b\u5132\u5b58\u5931\u6557',
+ settings_aux_saved: '\u8f14\u52a9\u6a21\u578b\u5df2\u66f4\u65b0',
+ settings_aux_no_changes: '\u6c92\u6709\u9700\u8981\u61c9\u7528\u7684\u8b8a\u66f4',
settings_label_send_key: '\u767c\u9001\u5feb\u6377\u9375',
settings_label_theme: '\u4e3b\u984c',
settings_label_skin: '佈景',
@@ -10096,6 +10224,22 @@ const LOCALES = {
settings_save_btn: 'Salvar Configurações',
settings_label_model: 'Modelo Padrão',
settings_desc_model: 'Usado para novas conversas. Conversas existentes mantêm o modelo selecionado.',
+ settings_label_auxiliary_models: 'Modelos auxiliares',
+ settings_desc_auxiliary_models: 'Roteamento para tarefas secundárias como visão, compressão, geração de títulos, etc. "Auto" usa o modelo de chat principal.',
+ settings_btn_reset_aux_models: 'Restaurar tudo para auto',
+ settings_btn_apply_aux_models: 'Aplicar alterações',
+ settings_aux_provider_auto: 'usar modelo principal',
+ settings_aux_model_auto: 'auto (usar padrão do provedor)',
+ settings_aux_model_custom: 'Modelo personalizado…',
+ settings_aux_model_custom_prompt: 'Insira o ID do modelo:',
+ settings_aux_loading: 'Carregando modelos auxiliares…',
+ settings_aux_load_failed: 'Não foi possível carregar as configurações de modelos auxiliares. Certifique-se de que a API do agente esteja disponível.',
+ settings_aux_reset_confirm_title: 'Restaurar modelos auxiliares?',
+ settings_aux_reset_confirm_msg: 'Isso definirá todas as tarefas auxiliares para auto (usar modelo principal).',
+ settings_aux_reset_done: 'Modelos auxiliares restaurados para auto',
+ settings_aux_save_failed: 'Falha ao salvar o modelo auxiliar',
+ settings_aux_saved: 'Modelos auxiliares atualizados',
+ settings_aux_no_changes: 'Nenhuma alteração para aplicar',
settings_label_send_key: 'Tecla de Envio',
settings_label_theme: 'Tema',
settings_label_skin: 'Skin',
@@ -11234,6 +11378,22 @@ const LOCALES = {
settings_save_btn: '설정 저장',
settings_label_model: '기본 모델',
settings_desc_model: '새 대화에 사용됩니다. 기존 대화는 선택된 모델을 유지합니다.',
+ settings_label_auxiliary_models: '보조 모델',
+ settings_desc_auxiliary_models: '비전, 압축, 제목 생성 등 보조 작업 라우팅. "자동"은 기본 채팅 모델을 사용합니다.',
+ settings_btn_reset_aux_models: '모두 자동으로 재설정',
+ settings_btn_apply_aux_models: '변경 사항 적용',
+ settings_aux_provider_auto: '기본 모델 사용',
+ settings_aux_model_auto: '자동 (제공자 기본값 사용)',
+ settings_aux_model_custom: '사용자 정의 모델…',
+ settings_aux_model_custom_prompt: '모델 ID 입력:',
+ settings_aux_loading: '보조 모델 로딩 중…',
+ settings_aux_load_failed: '보조 모델 설정을 로드할 수 없습니다. 에이전트 API가 사용 가능한지 확인하세요.',
+ settings_aux_reset_confirm_title: '보조 모델을 재설정하시겠습니까?',
+ settings_aux_reset_confirm_msg: '모든 보조 작업이 자동(기본 모델 사용)으로 설정됩니다.',
+ settings_aux_reset_done: '보조 모델이 자동으로 재설정됨',
+ settings_aux_save_failed: '보조 모델 저장 실패',
+ settings_aux_saved: '보조 모델 업데이트됨',
+ settings_aux_no_changes: '적용할 변경 사항 없음',
settings_label_send_key: '전송 키',
settings_label_theme: '테마',
settings_label_skin: '스킨',
@@ -12385,6 +12545,22 @@ const LOCALES = {
settings_save_btn: 'Enregistrer les paramètres',
settings_label_model: 'Modèle par défaut',
settings_desc_model: 'Utilisé pour les nouvelles conversations. Les conversations existantes conservent leur modèle sélectionné.',
+ settings_label_auxiliary_models: 'Modèles auxiliaires',
+ settings_desc_auxiliary_models: 'Routage des tâches secondaires : vision, compression, génération de titres, etc. « Auto » utilise le modèle de chat principal.',
+ settings_btn_reset_aux_models: 'Tout réinitialiser à auto',
+ settings_btn_apply_aux_models: 'Appliquer les modifications',
+ settings_aux_provider_auto: 'utiliser le modèle principal',
+ settings_aux_model_auto: 'auto (utiliser la valeur par défaut du fournisseur)',
+ settings_aux_model_custom: 'Modèle personnalisé…',
+ settings_aux_model_custom_prompt: 'Entrez l\u2019ID du modèle :',
+ settings_aux_loading: 'Chargement des modèles auxiliaires…',
+ settings_aux_load_failed: 'Impossible de charger les paramètres des modèles auxiliaires. Vérifiez que l\u2019API de l\u2019agent est disponible.',
+ settings_aux_reset_confirm_title: 'Réinitialiser les modèles auxiliaires ?',
+ settings_aux_reset_confirm_msg: 'Cela définira toutes les tâches auxiliaires sur auto (utiliser le modèle principal).',
+ settings_aux_reset_done: 'Modèles auxiliaires réinitialisés à auto',
+ settings_aux_save_failed: 'Échec de la sauvegarde du modèle auxiliaire',
+ settings_aux_saved: 'Modèles auxiliaires mis à jour',
+ settings_aux_no_changes: 'Aucune modification à appliquer',
settings_label_send_key: 'Envoyer la clé',
settings_label_theme: 'Thème',
settings_label_skin: 'Peau',
@@ -13637,7 +13813,23 @@ const LOCALES = {
settings_tab_system: 'Sistem',
settings_title: 'Ayarlar',
settings_save_btn: 'Ayarları Kaydet',
- settings_label_model: 'Varsayılan Model',
+ settings_label_model: 'Varsayılan Model',
+ settings_label_auxiliary_models: 'Yardımcı Modeller',
+ settings_desc_auxiliary_models: 'Görü, sıkıştırma, başlık oluşturma vb. yan görevler için yönlendirme. "Otomatik" ana sohbet modelinizi kullanır.',
+ settings_aux_loading: 'Yardımcı modeller yükleniyor…',
+ settings_aux_load_failed: 'Yardımcı model ayarları yüklenemedi. Agent API\'sinin kullanılabilir olduğundan emin olun.',
+ settings_aux_model_auto: 'otomatik (sağlayıcı varsayılanını kullan)',
+ settings_aux_provider_auto: 'ana modeli kullan',
+ settings_aux_model_custom: 'Özel model…',
+ settings_aux_model_custom_prompt: 'Model kimliğini girin:',
+ settings_aux_no_changes: 'Uygulanacak değişiklik yok',
+ settings_aux_saved: 'Yardımcı modeller güncellendi',
+ settings_aux_save_failed: 'Yardımcı model kaydedilemedi',
+ settings_aux_reset_confirm_title: 'Yardımcı modeller sıfırlansın mı?',
+ settings_aux_reset_confirm_msg: 'Bu, tüm yardımcı görevleri otomatik (ana modeli kullan) olarak ayarlayacaktır.',
+ settings_aux_reset_done: 'Yardımcı modeller otomatiğe sıfırlandı',
+ settings_btn_apply_aux_models: 'Değişiklikleri uygula',
+ settings_btn_reset_aux_models: 'Tümünü otomatiğe sıfırla',
settings_desc_model: 'Yeni konuşmalar için kullanılır. Mevcut konuşmalar seçilen modellerini korur.',
settings_label_send_key: 'Anahtar Gönder',
settings_label_theme: 'Tema',
diff --git a/static/index.html b/static/index.html
index e5518943..222998e8 100644
--- a/static/index.html
+++ b/static/index.html
@@ -2,7 +2,7 @@
-
+
Hermes