Merge branch 'master-nj' into 'master'
feat: 添加管理员模型配置页面与 config.yaml 模型管理 API See merge request knowledge-base-bj/deerflow-code!19
This commit is contained in:
董宏凯 committed
commit
2a869f4ff6
20 files changed
+1720
-179
No files matched your search
+5
-2
@@ -80,8 +80,11 @@ offline-backend-20260512/backend/skill_evolution.json
|
||||
skills/
|
||||
hermes-agent-main/
|
||||
.gitignore
|
||||
offline-backend-20260512/backend/tests/*
|
||||
deploy
|
||||
offline-backend-20260512/backend/tests/*
|
||||
# Keep the model configuration regression coverage versioned even though
|
||||
# this deployment repository otherwise ignores newly added backend tests.
|
||||
!offline-backend-20260512/backend/tests/test_model_management.py
|
||||
deploy
|
||||
.history
|
||||
|
||||
# Cloned third-party repos (local only)
|
||||
|
||||
@@ -2,6 +2,7 @@ import { useEffect, useState } from "react";
|
||||
import { useNavigate, useLocation } from "react-router-dom";
|
||||
import {
|
||||
Bell,
|
||||
Cpu,
|
||||
ChevronLeft,
|
||||
ChevronRight,
|
||||
ChevronsUpDown,
|
||||
@@ -252,6 +253,14 @@ function SecondaryNavSidebar({ headerHidden }: { headerHidden: boolean }) {
|
||||
白名单管理
|
||||
</DropdownMenuItem>
|
||||
)}
|
||||
{isAdmin && (
|
||||
<DropdownMenuItem
|
||||
onClick={() => navigate("/page/workspace/admin/models")}
|
||||
>
|
||||
<Cpu className="size-4" />
|
||||
模型配置
|
||||
</DropdownMenuItem>
|
||||
)}
|
||||
{isAdmin && (
|
||||
<DropdownMenuItem
|
||||
onClick={() => navigate("/page/strategy/settings/model-concurrency")}
|
||||
@@ -555,6 +564,12 @@ function LegacyAccordionSidebar({
|
||||
岗位角色配置
|
||||
</DropdownMenuItem>
|
||||
)}
|
||||
{isAdmin && (
|
||||
<DropdownMenuItem onClick={() => navigate("/page/workspace/admin/models")}>
|
||||
<Cpu className="size-4" />
|
||||
模型配置
|
||||
</DropdownMenuItem>
|
||||
)}
|
||||
{isAdmin && (
|
||||
<DropdownMenuItem onClick={() => navigate("/page/strategy/admin/login-whitelist")}>
|
||||
<ShieldCheck className="size-4" />
|
||||
|
||||
@@ -15,6 +15,7 @@ import { Tabs, TabsContent, TabsList, TabsTrigger } from "@/components/ui/tabs";
|
||||
import { apiFetch } from "@/core/api";
|
||||
import { getBackendBaseURL } from "@/core/config";
|
||||
import { isThreadRagModeLocked, useLocalSettings } from "@/core/settings";
|
||||
import { isSingleRoundRagMessage } from "@/core/threads/use-single-round-rag";
|
||||
import { useSkills } from "@/core/skills/hooks";
|
||||
import type { SkillDisplayConfig } from "@/core/skills/type";
|
||||
import {
|
||||
@@ -92,6 +93,13 @@ const ChatBox: React.FC<{
|
||||
const [settings] = useLocalSettings();
|
||||
const ragModeEnabled =
|
||||
settings.display.ragModeEnabled || isThreadRagModeLocked(threadId);
|
||||
// Frontend single-round RAG writes its own source tool message, so it should
|
||||
// use the same reference panel without changing the user's global RAG mode.
|
||||
const hasSingleRoundRag = useMemo(
|
||||
() => (thread.values.messages ?? []).some(isSingleRoundRagMessage),
|
||||
[thread.values.messages],
|
||||
);
|
||||
const referencesEnabled = ragModeEnabled || hasSingleRoundRag;
|
||||
const { skills } = useSkills();
|
||||
|
||||
const {
|
||||
@@ -106,11 +114,11 @@ const ChatBox: React.FC<{
|
||||
|
||||
const [autoSelectFirstArtifact, setAutoSelectFirstArtifact] = useState(true);
|
||||
const skillDisplayMap = useMemo(
|
||||
() => (ragModeEnabled ? buildSkillDisplayMap(skills) : new Map()),
|
||||
[ragModeEnabled, skills],
|
||||
() => (referencesEnabled ? buildSkillDisplayMap(skills) : new Map()),
|
||||
[referencesEnabled, skills],
|
||||
);
|
||||
const fallbackReferenceBatches = useMemo(() => {
|
||||
if (!ragModeEnabled) {
|
||||
if (!referencesEnabled) {
|
||||
return [];
|
||||
}
|
||||
return buildTurnReferenceBatches(
|
||||
@@ -118,9 +126,9 @@ const ChatBox: React.FC<{
|
||||
skillDisplayMap,
|
||||
thread.isLoading,
|
||||
);
|
||||
}, [ragModeEnabled, skillDisplayMap, thread.isLoading, thread.values.messages]);
|
||||
}, [referencesEnabled, skillDisplayMap, thread.isLoading, thread.values.messages]);
|
||||
useEffect(() => {
|
||||
if (!ragModeEnabled) {
|
||||
if (!referencesEnabled) {
|
||||
setBackendReferenceBatches([]);
|
||||
return;
|
||||
}
|
||||
@@ -156,7 +164,7 @@ const ChatBox: React.FC<{
|
||||
}
|
||||
});
|
||||
return () => controller.abort();
|
||||
}, [ragModeEnabled, thread.isLoading, threadId, thread.values.messages]);
|
||||
}, [referencesEnabled, thread.isLoading, threadId, thread.values.messages]);
|
||||
const adaptedBackendReferenceBatches = useMemo(
|
||||
() =>
|
||||
adaptBackendReferenceBatches(
|
||||
@@ -175,7 +183,7 @@ const ChatBox: React.FC<{
|
||||
[adaptedBackendReferenceBatches, fallbackReferenceBatches],
|
||||
);
|
||||
useEffect(() => {
|
||||
if (!ragModeEnabled) {
|
||||
if (!referencesEnabled) {
|
||||
return;
|
||||
}
|
||||
const signature = JSON.stringify({
|
||||
@@ -202,7 +210,7 @@ const ChatBox: React.FC<{
|
||||
fallbackBatches: debugReferenceBatches(fallbackReferenceBatches),
|
||||
finalBatches: debugReferenceBatches(referenceBatches),
|
||||
});
|
||||
}, [adaptedBackendReferenceBatches, backendReferenceBatches, fallbackReferenceBatches, ragModeEnabled, referenceBatches, threadId]);
|
||||
}, [adaptedBackendReferenceBatches, backendReferenceBatches, fallbackReferenceBatches, referencesEnabled, referenceBatches, threadId]);
|
||||
const latestReferenceBatch =
|
||||
referenceBatches.length > 0
|
||||
? referenceBatches[referenceBatches.length - 1]!
|
||||
@@ -213,7 +221,7 @@ const ChatBox: React.FC<{
|
||||
const referenceSources = activeReferenceBatch?.sources ?? [];
|
||||
const hasCitationReferences = referenceSources.length > 0;
|
||||
useEffect(() => {
|
||||
if (!ragModeEnabled || !activeReferenceBatch) {
|
||||
if (!referencesEnabled || !activeReferenceBatch) {
|
||||
return;
|
||||
}
|
||||
logReferenceDebug("frontend-active-reference-batch-payload", threadId, {
|
||||
@@ -222,7 +230,7 @@ const ChatBox: React.FC<{
|
||||
sourceCount: activeReferenceBatch.sources.length,
|
||||
sources: activeReferenceBatch.sources.map(debugReferenceSource),
|
||||
});
|
||||
}, [activeReferenceBatch, ragModeEnabled, threadId]);
|
||||
}, [activeReferenceBatch, referencesEnabled, threadId]);
|
||||
useEffect(() => {
|
||||
if (hasCitationReferences) {
|
||||
setReferencesOpen(true);
|
||||
|
||||
@@ -64,6 +64,7 @@ import { getBackendBaseURL } from "@/core/config";
|
||||
import { useI18n } from "@/core/i18n/hooks";
|
||||
import { isHiddenFromUIMessage } from "@/core/messages/utils";
|
||||
import { useModels } from "@/core/models/hooks";
|
||||
import type { Model } from "@/core/models/types";
|
||||
import type { AgentThreadContext } from "@/core/threads";
|
||||
import { textOfMessage } from "@/core/threads/utils";
|
||||
import { desensitize } from "@/lib/desensitize";
|
||||
@@ -133,8 +134,9 @@ export function InputBox({
|
||||
onSubmit,
|
||||
onStop,
|
||||
promptPrefix,
|
||||
agentSelector,
|
||||
assistantId,
|
||||
agentSelector,
|
||||
extraModels = [],
|
||||
assistantId,
|
||||
hideAdminRecommendedQuestions,
|
||||
compact,
|
||||
...props
|
||||
@@ -190,16 +192,23 @@ export function InputBox({
|
||||
* homepage (``ChatPage``) passes this; ``AgentChatPage`` omits it because the
|
||||
* thread is already bound to one agent.
|
||||
*/
|
||||
agentSelector?: {
|
||||
selectedAgentId: string | null;
|
||||
onAgentChange: (agentId: string | null) => void;
|
||||
};
|
||||
}) {
|
||||
agentSelector?: {
|
||||
selectedAgentId: string | null;
|
||||
onAgentChange: (agentId: string | null) => void;
|
||||
};
|
||||
/** Page-scoped model choices implemented outside the DeerFlow model API. */
|
||||
extraModels?: Model[];
|
||||
}) {
|
||||
const { t } = useI18n();
|
||||
const [localSettings, setLocalSettings] = useLocalSettings();
|
||||
const searchParams = useSearchParams();
|
||||
const [modelDialogOpen, setModelDialogOpen] = useState(false);
|
||||
const { models } = useModels();
|
||||
const [modelDialogOpen, setModelDialogOpen] = useState(false);
|
||||
const { models } = useModels();
|
||||
const singleRoundRagActive =
|
||||
context.execution_profile === "single_round_rag";
|
||||
const modelOptions = useMemo(() => {
|
||||
return singleRoundRagActive ? extraModels : models;
|
||||
}, [extraModels, models, singleRoundRagActive]);
|
||||
const { thread, isMock } = useThread();
|
||||
const { textInput } = usePromptInputController();
|
||||
const promptRootRef = useRef<HTMLDivElement | null>(null);
|
||||
@@ -223,34 +232,40 @@ export function InputBox({
|
||||
recommendedQuestions.length > 0;
|
||||
|
||||
useEffect(() => {
|
||||
if (models.length === 0) {
|
||||
return;
|
||||
}
|
||||
const currentModel = models.find((m) => m.name === context.model_name);
|
||||
const fallbackModel = currentModel ?? models[0]!;
|
||||
if (modelOptions.length === 0) {
|
||||
return;
|
||||
}
|
||||
const currentModel = modelOptions.find((m) => m.name === context.model_name);
|
||||
const fallbackModel = currentModel ?? modelOptions[0]!;
|
||||
const supportsThinking = fallbackModel.supports_thinking ?? false;
|
||||
const nextModelName = fallbackModel.name;
|
||||
const nextMode = getResolvedMode(context.mode, supportsThinking);
|
||||
|
||||
if (context.model_name === nextModelName && context.mode === nextMode) {
|
||||
return;
|
||||
}
|
||||
if (
|
||||
context.model_name === nextModelName &&
|
||||
context.mode === nextMode &&
|
||||
context.execution_profile === fallbackModel.execution_profile
|
||||
) {
|
||||
return;
|
||||
}
|
||||
|
||||
onContextChange?.({
|
||||
...context,
|
||||
model_name: nextModelName,
|
||||
mode: nextMode,
|
||||
});
|
||||
}, [context, models, onContextChange]);
|
||||
...context,
|
||||
model_name: nextModelName,
|
||||
mode: nextMode,
|
||||
execution_profile: fallbackModel.execution_profile,
|
||||
});
|
||||
}, [context, modelOptions, onContextChange]);
|
||||
|
||||
const selectedModel = useMemo(() => {
|
||||
if (modelOptions.length === 0) {
|
||||
return undefined;
|
||||
}
|
||||
return modelOptions.find((m) => m.name === context.model_name) ?? modelOptions[0];
|
||||
}, [context.model_name, modelOptions]);
|
||||
|
||||
const selectedModel = useMemo(() => {
|
||||
if (models.length === 0) {
|
||||
return undefined;
|
||||
}
|
||||
return models.find((m) => m.name === context.model_name) ?? models[0];
|
||||
}, [context.model_name, models]);
|
||||
|
||||
const resolvedModelName = selectedModel?.name;
|
||||
const resolvedModelName = selectedModel?.name;
|
||||
const selectedSingleRoundRag = singleRoundRagActive;
|
||||
|
||||
const supportThinking = useMemo(
|
||||
() => selectedModel?.supports_thinking ?? false,
|
||||
@@ -264,38 +279,65 @@ export function InputBox({
|
||||
|
||||
const handleModelSelect = useCallback(
|
||||
(model_name: string) => {
|
||||
const model = models.find((m) => m.name === model_name);
|
||||
const model = modelOptions.find((m) => m.name === model_name);
|
||||
if (!model) {
|
||||
return;
|
||||
}
|
||||
onContextChange?.({
|
||||
...context,
|
||||
model_name,
|
||||
mode: getResolvedMode(context.mode, model.supports_thinking ?? false),
|
||||
reasoning_effort: context.reasoning_effort,
|
||||
});
|
||||
setModelDialogOpen(false);
|
||||
},
|
||||
[onContextChange, context, models],
|
||||
mode: getResolvedMode(context.mode, model.supports_thinking ?? false),
|
||||
reasoning_effort: context.reasoning_effort,
|
||||
execution_profile: model.execution_profile,
|
||||
});
|
||||
setModelDialogOpen(false);
|
||||
},
|
||||
[onContextChange, context, modelOptions],
|
||||
);
|
||||
|
||||
const handleModeSelect = useCallback(
|
||||
(mode: InputMode) => {
|
||||
onContextChange?.({
|
||||
...context,
|
||||
mode: getResolvedMode(mode, supportThinking),
|
||||
reasoning_effort:
|
||||
mode === "ultra"
|
||||
? "high"
|
||||
const handleModeSelect = useCallback(
|
||||
(mode: InputMode) => {
|
||||
const ordinaryModel =
|
||||
models.find((model) => model.name === context.model_name) ?? models[0];
|
||||
onContextChange?.({
|
||||
...context,
|
||||
model_name: ordinaryModel?.name ?? context.model_name,
|
||||
mode: getResolvedMode(
|
||||
mode,
|
||||
ordinaryModel?.supports_thinking ?? supportThinking,
|
||||
),
|
||||
reasoning_effort:
|
||||
mode === "ultra"
|
||||
? "high"
|
||||
: mode === "pro"
|
||||
? "medium"
|
||||
: mode === "thinking"
|
||||
? "low"
|
||||
: "minimal",
|
||||
});
|
||||
},
|
||||
[onContextChange, context, supportThinking],
|
||||
);
|
||||
? "low"
|
||||
: "minimal",
|
||||
execution_profile: undefined,
|
||||
});
|
||||
},
|
||||
[onContextChange, context, models, supportThinking],
|
||||
);
|
||||
|
||||
const handleSingleRoundRagSelect = useCallback(() => {
|
||||
const ragModel =
|
||||
extraModels.find((model) => model.name === context.model_name) ??
|
||||
extraModels[0];
|
||||
if (!ragModel) {
|
||||
return;
|
||||
}
|
||||
|
||||
onContextChange?.({
|
||||
...context,
|
||||
model_name: ragModel.name,
|
||||
mode: getResolvedMode(
|
||||
context.mode,
|
||||
ragModel.supports_thinking ?? false,
|
||||
),
|
||||
execution_profile: "single_round_rag",
|
||||
});
|
||||
}, [context, extraModels, onContextChange]);
|
||||
|
||||
const handleReasoningEffortSelect = useCallback(
|
||||
(effort: "minimal" | "low" | "medium" | "high") => {
|
||||
@@ -361,10 +403,11 @@ export function InputBox({
|
||||
onContextChange?.({
|
||||
...context,
|
||||
model_name: resolvedModelName,
|
||||
mode: getResolvedMode(
|
||||
context.mode,
|
||||
selectedModel?.supports_thinking ?? false,
|
||||
),
|
||||
mode: getResolvedMode(
|
||||
context.mode,
|
||||
selectedModel?.supports_thinking ?? false,
|
||||
),
|
||||
execution_profile: selectedModel?.execution_profile,
|
||||
});
|
||||
setTimeout(() => onSubmit?.(message), 0);
|
||||
return;
|
||||
@@ -379,7 +422,8 @@ export function InputBox({
|
||||
onStop,
|
||||
resolvedModelName,
|
||||
ragModeEnabled,
|
||||
selectedModel?.supports_thinking,
|
||||
selectedModel?.supports_thinking,
|
||||
selectedModel?.execution_profile,
|
||||
status,
|
||||
threadId,
|
||||
],
|
||||
@@ -437,8 +481,9 @@ export function InputBox({
|
||||
}, [pendingSuggestion, requestFormSubmit, textInput]);
|
||||
|
||||
const showFollowups =
|
||||
!disabled &&
|
||||
!isNewThread &&
|
||||
!disabled &&
|
||||
!isNewThread &&
|
||||
!selectedSingleRoundRag &&
|
||||
!followupsHidden &&
|
||||
(followupsLoading || followups.length > 0);
|
||||
|
||||
@@ -464,7 +509,7 @@ export function InputBox({
|
||||
return;
|
||||
}
|
||||
|
||||
if (disabled || isMock) {
|
||||
if (disabled || isMock || selectedSingleRoundRag) {
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -526,7 +571,7 @@ export function InputBox({
|
||||
});
|
||||
|
||||
return () => controller.abort();
|
||||
}, [context.model_name, disabled, isMock, status, thread.messages, threadId]);
|
||||
}, [context.model_name, disabled, isMock, selectedSingleRoundRag, status, thread.messages, threadId]);
|
||||
|
||||
return (
|
||||
<div ref={promptRootRef} className="relative flex flex-col gap-6">
|
||||
@@ -782,54 +827,64 @@ export function InputBox({
|
||||
</PromptInputButton>
|
||||
</Tooltip>
|
||||
<PromptInputActionMenu>
|
||||
<ModeHoverGuide
|
||||
mode={
|
||||
<ModeHoverGuide
|
||||
mode={
|
||||
context.mode === "flash" ||
|
||||
context.mode === "thinking" ||
|
||||
context.mode === "pro" ||
|
||||
context.mode === "ultra"
|
||||
? context.mode
|
||||
: "flash"
|
||||
}
|
||||
>
|
||||
? context.mode
|
||||
: "flash"
|
||||
}
|
||||
tooltipContent={
|
||||
selectedSingleRoundRag
|
||||
? "单轮 RAG 问答：基于知识库检索生成答案，并展示参考资料"
|
||||
: undefined
|
||||
}
|
||||
>
|
||||
<PromptInputActionMenuTrigger
|
||||
className={cn(
|
||||
"gap-1.5! rounded-full border-0 bg-transparent px-3! text-muted-foreground shadow-none hover:bg-muted/70 hover:text-foreground dark:text-slate-400 dark:hover:bg-slate-700/70 dark:hover:text-slate-100",
|
||||
compact && "rounded-lg border px-2.5! py-1! text-[11px] font-medium",
|
||||
compact &&
|
||||
(context.mode === "thinking"
|
||||
? "border-indigo-200 bg-indigo-50/60 text-indigo-700 hover:bg-indigo-100 dark:border-indigo-700 dark:bg-indigo-950/40 dark:text-indigo-200"
|
||||
compact &&
|
||||
(selectedSingleRoundRag
|
||||
? "border-emerald-200 bg-emerald-50/60 text-emerald-700 hover:bg-emerald-100 dark:border-emerald-700 dark:bg-emerald-950/40 dark:text-emerald-200"
|
||||
: context.mode === "thinking"
|
||||
? "border-indigo-200 bg-indigo-50/60 text-indigo-700 hover:bg-indigo-100 dark:border-indigo-700 dark:bg-indigo-950/40 dark:text-indigo-200"
|
||||
: context.mode === "pro"
|
||||
? "border-violet-200 bg-violet-50/60 text-violet-700 hover:bg-violet-100 dark:border-violet-700 dark:bg-violet-950/40 dark:text-violet-200"
|
||||
: context.mode === "ultra"
|
||||
? "border-amber-200 bg-amber-50/60 text-amber-700 hover:bg-amber-100 dark:border-amber-700 dark:bg-amber-950/40 dark:text-amber-200"
|
||||
: "border-transparent bg-muted/40 hover:bg-muted hover:text-foreground"),
|
||||
)}
|
||||
>
|
||||
<div>
|
||||
{context.mode === "flash" && <ZapIcon className="size-3" />}
|
||||
{context.mode === "thinking" && (
|
||||
<LightbulbIcon className="size-3" />
|
||||
)}
|
||||
{context.mode === "pro" && (
|
||||
<GraduationCapIcon className="size-3" />
|
||||
)}
|
||||
{context.mode === "ultra" && (
|
||||
<RocketIcon className="size-3 text-[#dabb5e]" />
|
||||
)}
|
||||
</div>
|
||||
>
|
||||
<div>
|
||||
{selectedSingleRoundRag ? (
|
||||
<BookMarkedIcon className="size-3" />
|
||||
) : context.mode === "flash" ? (
|
||||
<ZapIcon className="size-3" />
|
||||
) : context.mode === "thinking" ? (
|
||||
<LightbulbIcon className="size-3" />
|
||||
) : context.mode === "pro" ? (
|
||||
<GraduationCapIcon className="size-3" />
|
||||
) : context.mode === "ultra" ? (
|
||||
<RocketIcon className="size-3 text-[#dabb5e]" />
|
||||
) : null}
|
||||
</div>
|
||||
<div
|
||||
className={cn(
|
||||
"text-xs font-normal",
|
||||
context.mode === "ultra" ? "golden-text" : "",
|
||||
)}
|
||||
>
|
||||
{(context.mode === "flash" && t.inputBox.flashMode) ||
|
||||
(context.mode === "thinking" &&
|
||||
t.inputBox.reasoningMode) ||
|
||||
(context.mode === "pro" && t.inputBox.proMode) ||
|
||||
(context.mode === "ultra" && t.inputBox.ultraMode)}
|
||||
</div>
|
||||
{selectedSingleRoundRag
|
||||
? "单轮 RAG"
|
||||
: (context.mode === "flash" && t.inputBox.flashMode) ||
|
||||
(context.mode === "thinking" &&
|
||||
t.inputBox.reasoningMode) ||
|
||||
(context.mode === "pro" && t.inputBox.proMode) ||
|
||||
(context.mode === "ultra" && t.inputBox.ultraMode)}
|
||||
</div>
|
||||
</PromptInputActionMenuTrigger>
|
||||
</ModeHoverGuide>
|
||||
<PromptInputActionMenuContent className="w-80">
|
||||
@@ -838,11 +893,11 @@ export function InputBox({
|
||||
{t.inputBox.mode}
|
||||
</DropdownMenuLabel>
|
||||
<PromptInputActionMenu>
|
||||
<PromptInputActionMenuItem
|
||||
className={cn(
|
||||
context.mode === "flash"
|
||||
? "text-accent-foreground"
|
||||
: "text-muted-foreground/65",
|
||||
<PromptInputActionMenuItem
|
||||
className={cn(
|
||||
!selectedSingleRoundRag && context.mode === "flash"
|
||||
? "text-accent-foreground"
|
||||
: "text-muted-foreground/65",
|
||||
)}
|
||||
onSelect={() => handleModeSelect("flash")}
|
||||
>
|
||||
@@ -850,9 +905,9 @@ export function InputBox({
|
||||
<div className="flex items-center gap-1 font-bold">
|
||||
<ZapIcon
|
||||
className={cn(
|
||||
"mr-2 size-4",
|
||||
context.mode === "flash" &&
|
||||
"text-accent-foreground",
|
||||
"mr-2 size-4",
|
||||
!selectedSingleRoundRag && context.mode === "flash" &&
|
||||
"text-accent-foreground",
|
||||
)}
|
||||
/>
|
||||
{t.inputBox.flashMode}
|
||||
@@ -861,7 +916,7 @@ export function InputBox({
|
||||
{t.inputBox.flashModeDescription}
|
||||
</div>
|
||||
</div>
|
||||
{context.mode === "flash" ? (
|
||||
{!selectedSingleRoundRag && context.mode === "flash" ? (
|
||||
<CheckIcon className="ml-auto size-4" />
|
||||
) : (
|
||||
<div className="ml-auto size-4" />
|
||||
@@ -869,9 +924,9 @@ export function InputBox({
|
||||
</PromptInputActionMenuItem>
|
||||
{supportThinking && (
|
||||
<PromptInputActionMenuItem
|
||||
className={cn(
|
||||
context.mode === "thinking"
|
||||
? "text-accent-foreground"
|
||||
className={cn(
|
||||
!selectedSingleRoundRag && context.mode === "thinking"
|
||||
? "text-accent-foreground"
|
||||
: "text-muted-foreground/65",
|
||||
)}
|
||||
onSelect={() => handleModeSelect("thinking")}
|
||||
@@ -880,9 +935,9 @@ export function InputBox({
|
||||
<div className="flex items-center gap-1 font-bold">
|
||||
<LightbulbIcon
|
||||
className={cn(
|
||||
"mr-2 size-4",
|
||||
context.mode === "thinking" &&
|
||||
"text-accent-foreground",
|
||||
"mr-2 size-4",
|
||||
!selectedSingleRoundRag && context.mode === "thinking" &&
|
||||
"text-accent-foreground",
|
||||
)}
|
||||
/>
|
||||
{t.inputBox.reasoningMode}
|
||||
@@ -891,7 +946,7 @@ export function InputBox({
|
||||
{t.inputBox.reasoningModeDescription}
|
||||
</div>
|
||||
</div>
|
||||
{context.mode === "thinking" ? (
|
||||
{!selectedSingleRoundRag && context.mode === "thinking" ? (
|
||||
<CheckIcon className="ml-auto size-4" />
|
||||
) : (
|
||||
<div className="ml-auto size-4" />
|
||||
@@ -899,8 +954,8 @@ export function InputBox({
|
||||
</PromptInputActionMenuItem>
|
||||
)}
|
||||
<PromptInputActionMenuItem
|
||||
className={cn(
|
||||
context.mode === "pro"
|
||||
className={cn(
|
||||
!selectedSingleRoundRag && context.mode === "pro"
|
||||
? "text-accent-foreground"
|
||||
: "text-muted-foreground/65",
|
||||
)}
|
||||
@@ -910,8 +965,8 @@ export function InputBox({
|
||||
<div className="flex items-center gap-1 font-bold">
|
||||
<GraduationCapIcon
|
||||
className={cn(
|
||||
"mr-2 size-4",
|
||||
context.mode === "pro" &&
|
||||
"mr-2 size-4",
|
||||
!selectedSingleRoundRag && context.mode === "pro" &&
|
||||
"text-accent-foreground",
|
||||
)}
|
||||
/>
|
||||
@@ -921,15 +976,15 @@ export function InputBox({
|
||||
{t.inputBox.proModeDescription}
|
||||
</div>
|
||||
</div>
|
||||
{context.mode === "pro" ? (
|
||||
{!selectedSingleRoundRag && context.mode === "pro" ? (
|
||||
<CheckIcon className="ml-auto size-4" />
|
||||
) : (
|
||||
<div className="ml-auto size-4" />
|
||||
)}
|
||||
</PromptInputActionMenuItem>
|
||||
<PromptInputActionMenuItem
|
||||
className={cn(
|
||||
context.mode === "ultra"
|
||||
className={cn(
|
||||
!selectedSingleRoundRag && context.mode === "ultra"
|
||||
? "text-accent-foreground"
|
||||
: "text-muted-foreground/65",
|
||||
)}
|
||||
@@ -939,13 +994,17 @@ export function InputBox({
|
||||
<div className="flex items-center gap-1 font-bold">
|
||||
<RocketIcon
|
||||
className={cn(
|
||||
"mr-2 size-4",
|
||||
context.mode === "ultra" && "text-[#dabb5e]",
|
||||
"mr-2 size-4",
|
||||
!selectedSingleRoundRag &&
|
||||
context.mode === "ultra" &&
|
||||
"text-[#dabb5e]",
|
||||
)}
|
||||
/>
|
||||
<div
|
||||
className={cn(
|
||||
context.mode === "ultra" && "golden-text",
|
||||
!selectedSingleRoundRag &&
|
||||
context.mode === "ultra" &&
|
||||
"golden-text",
|
||||
)}
|
||||
>
|
||||
{t.inputBox.ultraMode}
|
||||
@@ -955,13 +1014,46 @@ export function InputBox({
|
||||
{t.inputBox.ultraModeDescription}
|
||||
</div>
|
||||
</div>
|
||||
{context.mode === "ultra" ? (
|
||||
{!selectedSingleRoundRag && context.mode === "ultra" ? (
|
||||
<CheckIcon className="ml-auto size-4" />
|
||||
) : (
|
||||
<div className="ml-auto size-4" />
|
||||
)}
|
||||
</PromptInputActionMenuItem>
|
||||
</PromptInputActionMenu>
|
||||
</PromptInputActionMenuItem>
|
||||
{extraModels.length > 0 && (
|
||||
<>
|
||||
<DropdownMenuSeparator className="my-2" />
|
||||
<PromptInputActionMenuItem
|
||||
className={cn(
|
||||
selectedSingleRoundRag
|
||||
? "bg-emerald-50 text-emerald-800 hover:bg-emerald-100 focus:bg-emerald-100 dark:bg-emerald-950/40 dark:text-emerald-200 dark:hover:bg-emerald-950/60 dark:focus:bg-emerald-950/60"
|
||||
: "text-muted-foreground/65",
|
||||
)}
|
||||
onSelect={handleSingleRoundRagSelect}
|
||||
>
|
||||
<div className="flex flex-col gap-2">
|
||||
<div className="flex items-center gap-1 font-bold">
|
||||
<BookMarkedIcon
|
||||
className={cn(
|
||||
"mr-2 size-4",
|
||||
selectedSingleRoundRag && "text-emerald-700 dark:text-emerald-200",
|
||||
)}
|
||||
/>
|
||||
单轮 RAG 问答
|
||||
</div>
|
||||
<div className="pl-7 text-xs">
|
||||
基于知识库检索生成答案，并展示思考内容与参考资料
|
||||
</div>
|
||||
</div>
|
||||
{selectedSingleRoundRag ? (
|
||||
<CheckIcon className="ml-auto size-4" />
|
||||
) : (
|
||||
<div className="ml-auto size-4" />
|
||||
)}
|
||||
</PromptInputActionMenuItem>
|
||||
</>
|
||||
)}
|
||||
</PromptInputActionMenu>
|
||||
</DropdownMenuGroup>
|
||||
</PromptInputActionMenuContent>
|
||||
</PromptInputActionMenu>
|
||||
@@ -1146,7 +1238,7 @@ export function InputBox({
|
||||
<ModelSelectorContent>
|
||||
<ModelSelectorInput placeholder={t.inputBox.searchModels} />
|
||||
<ModelSelectorList>
|
||||
{models.map((m) => (
|
||||
{modelOptions.map((m) => (
|
||||
<ModelSelectorItem
|
||||
key={m.name}
|
||||
value={m.name}
|
||||
@@ -1154,9 +1246,11 @@ export function InputBox({
|
||||
>
|
||||
<div className="flex min-w-0 flex-1 flex-col">
|
||||
<ModelSelectorName>{m.display_name}</ModelSelectorName>
|
||||
<span className="text-muted-foreground truncate text-[10px]">
|
||||
{m.model}
|
||||
</span>
|
||||
<span className="text-muted-foreground truncate text-[10px]">
|
||||
{selectedSingleRoundRag
|
||||
? `单轮知识库问答 · ${m.name}`
|
||||
: m.model}
|
||||
</span>
|
||||
</div>
|
||||
{m.name === context.model_name ? (
|
||||
<CheckIcon className="ml-auto size-4" />
|
||||
@@ -1164,8 +1258,8 @@ export function InputBox({
|
||||
<div className="ml-auto size-4" />
|
||||
)}
|
||||
</ModelSelectorItem>
|
||||
))}
|
||||
</ModelSelectorList>
|
||||
))}
|
||||
</ModelSelectorList>
|
||||
</ModelSelectorContent>
|
||||
</ModelSelector>
|
||||
<PromptInputSubmit
|
||||
|
||||
@@ -33,7 +33,8 @@ import {
|
||||
} from "@/core/skill-display/runtime";
|
||||
import type { Subtask } from "@/core/tasks";
|
||||
import { useUpdateSubtask } from "@/core/tasks/context";
|
||||
import type { AgentThreadState } from "@/core/threads";
|
||||
import type { AgentThreadState } from "@/core/threads";
|
||||
import { isSingleRoundRagMessage } from "@/core/threads/use-single-round-rag";
|
||||
import { cn } from "@/lib/utils";
|
||||
import { resolveAllowMultiple } from "@/roundtable-planning/lib/clarification";
|
||||
|
||||
@@ -235,12 +236,17 @@ export function MessageList({
|
||||
}, [messages, askTimes]);
|
||||
|
||||
const [settings] = useLocalSettings();
|
||||
const ragModeEnabled =
|
||||
settings.display.ragModeEnabled || isThreadRagModeLocked(threadId);
|
||||
const { skills } = useSkills();
|
||||
const skillDisplayMap = useMemo(
|
||||
() => (ragModeEnabled ? buildSkillDisplayMap(skills) : new Map()),
|
||||
[ragModeEnabled, skills],
|
||||
const ragModeEnabled =
|
||||
settings.display.ragModeEnabled || isThreadRagModeLocked(threadId);
|
||||
const hasSingleRoundRag = useMemo(
|
||||
() => messages.some(isSingleRoundRagMessage),
|
||||
[messages],
|
||||
);
|
||||
const referencesEnabled = ragModeEnabled || hasSingleRoundRag;
|
||||
const { skills } = useSkills();
|
||||
const skillDisplayMap = useMemo(
|
||||
() => (referencesEnabled ? buildSkillDisplayMap(skills) : new Map()),
|
||||
[referencesEnabled, skills],
|
||||
);
|
||||
// Same per-turn reference batches the right-side panel renders (backend
|
||||
// batches preferred). Inline [n] tokens must resolve against these so badge
|
||||
@@ -373,7 +379,7 @@ export function MessageList({
|
||||
if (group.type === "human" || group.type === "assistant") {
|
||||
return group.messages.map((msg) => {
|
||||
const isAssistantMessage =
|
||||
ragModeEnabled &&
|
||||
referencesEnabled &&
|
||||
group.type === "assistant" &&
|
||||
msg.type === "ai";
|
||||
const batchSources = isAssistantMessage
|
||||
|
||||
@@ -46,20 +46,23 @@ function getModeDescriptionKey(
|
||||
}
|
||||
}
|
||||
|
||||
export function ModeHoverGuide({
|
||||
mode,
|
||||
children,
|
||||
showTitle = true,
|
||||
}: {
|
||||
mode: AgentMode;
|
||||
children: React.ReactNode;
|
||||
/** When true, tooltip shows "ModeName: Description". When false, only description. */
|
||||
showTitle?: boolean;
|
||||
}) {
|
||||
const { t } = useI18n();
|
||||
const label = t.inputBox[getModeLabelKey(mode)];
|
||||
const description = t.inputBox[getModeDescriptionKey(mode)];
|
||||
const content = showTitle ? `${label}: ${description}` : description;
|
||||
export function ModeHoverGuide({
|
||||
mode,
|
||||
children,
|
||||
showTitle = true,
|
||||
tooltipContent,
|
||||
}: {
|
||||
mode: AgentMode;
|
||||
children: React.ReactNode;
|
||||
/** When true, tooltip shows "ModeName: Description". When false, only description. */
|
||||
showTitle?: boolean;
|
||||
/** Overrides the standard mode tooltip for page-specific execution modes. */
|
||||
tooltipContent?: string;
|
||||
}) {
|
||||
const { t } = useI18n();
|
||||
const label = t.inputBox[getModeLabelKey(mode)];
|
||||
const description = t.inputBox[getModeDescriptionKey(mode)];
|
||||
const content = tooltipContent ?? (showTitle ? `${label}: ${description}` : description);
|
||||
|
||||
return <Tooltip content={content}>{children}</Tooltip>;
|
||||
}
|
||||
@@ -184,7 +184,7 @@ export function WorkspaceNavChatList({
|
||||
</Link>
|
||||
</SidebarMenuButton>
|
||||
</SidebarMenuItem>
|
||||
<SidebarMenuItem>
|
||||
{/* <SidebarMenuItem>
|
||||
<SidebarMenuButton
|
||||
isActive={pathname.startsWith("/page/workspace/roundtable/position")}
|
||||
asChild
|
||||
@@ -195,7 +195,7 @@ export function WorkspaceNavChatList({
|
||||
<span>岗位会商</span>
|
||||
</Link>
|
||||
</SidebarMenuButton>
|
||||
</SidebarMenuItem>
|
||||
</SidebarMenuItem> */}
|
||||
<SidebarMenuItem>
|
||||
<SidebarMenuButton
|
||||
isActive={pathname === "/page/workspace/scheduled/tasks"}
|
||||
|
||||
@@ -0,0 +1,107 @@
|
||||
import { apiFetch } from "@/core/api";
|
||||
import { getBackendBaseURL } from "@/core/config";
|
||||
|
||||
export interface ManagedModel {
|
||||
name: string;
|
||||
display_name: string | null;
|
||||
description: string | null;
|
||||
use: string;
|
||||
model: string;
|
||||
base_url: string | null;
|
||||
request_timeout: number | null;
|
||||
max_retries: number | null;
|
||||
max_tokens: number | null;
|
||||
temperature: number | null;
|
||||
stream_usage: boolean | null;
|
||||
supports_thinking: boolean;
|
||||
supports_reasoning_effort: boolean;
|
||||
supports_vision: boolean;
|
||||
when_thinking_enabled: Record<string, unknown> | null;
|
||||
when_thinking_disabled: Record<string, unknown> | null;
|
||||
thinking: Record<string, unknown> | null;
|
||||
use_responses_api: boolean | null;
|
||||
output_version: string | null;
|
||||
extra_settings: Record<string, unknown>;
|
||||
api_key_configured: boolean;
|
||||
api_key_source: string | null;
|
||||
source: "config.yaml";
|
||||
}
|
||||
|
||||
export interface ManagedModelsResponse {
|
||||
models: ManagedModel[];
|
||||
config_path: string;
|
||||
updated_at: string | null;
|
||||
}
|
||||
|
||||
export interface ModelSavePayload {
|
||||
name: string;
|
||||
display_name?: string | null;
|
||||
description?: string | null;
|
||||
use: string;
|
||||
model: string;
|
||||
api_key?: string;
|
||||
clear_api_key?: boolean;
|
||||
base_url?: string | null;
|
||||
request_timeout?: number | null;
|
||||
max_retries?: number | null;
|
||||
max_tokens?: number | null;
|
||||
temperature?: number | null;
|
||||
stream_usage?: boolean | null;
|
||||
supports_thinking: boolean;
|
||||
supports_reasoning_effort: boolean;
|
||||
supports_vision: boolean;
|
||||
when_thinking_enabled?: Record<string, unknown> | null;
|
||||
when_thinking_disabled?: Record<string, unknown> | null;
|
||||
thinking?: Record<string, unknown> | null;
|
||||
use_responses_api?: boolean | null;
|
||||
output_version?: string | null;
|
||||
extra_settings: Record<string, unknown>;
|
||||
}
|
||||
|
||||
export interface ModelConnectionTestResult {
|
||||
success: boolean;
|
||||
latency_ms: number;
|
||||
message: string;
|
||||
preview?: string | null;
|
||||
}
|
||||
|
||||
async function readError(response: Response, fallback: string): Promise<never> {
|
||||
const payload = await response.json().catch(() => null) as { detail?: string } | null;
|
||||
throw new Error(payload?.detail || fallback);
|
||||
}
|
||||
|
||||
export async function loadManagedModels(): Promise<ManagedModelsResponse> {
|
||||
const response = await apiFetch(`${getBackendBaseURL()}/api/admin/models`);
|
||||
if (!response.ok) return readError(response, "加载模型配置失败");
|
||||
return response.json() as Promise<ManagedModelsResponse>;
|
||||
}
|
||||
|
||||
export async function createManagedModel(payload: ModelSavePayload): Promise<ManagedModel> {
|
||||
const response = await apiFetch(`${getBackendBaseURL()}/api/admin/models`, {
|
||||
method: "POST",
|
||||
headers: { "Content-Type": "application/json" },
|
||||
body: JSON.stringify(payload),
|
||||
});
|
||||
if (!response.ok) return readError(response, "创建模型失败");
|
||||
return response.json() as Promise<ManagedModel>;
|
||||
}
|
||||
|
||||
export async function updateManagedModel(name: string, payload: ModelSavePayload): Promise<ManagedModel> {
|
||||
const response = await apiFetch(`${getBackendBaseURL()}/api/admin/models/${encodeURIComponent(name)}`, {
|
||||
method: "PUT",
|
||||
headers: { "Content-Type": "application/json" },
|
||||
body: JSON.stringify(payload),
|
||||
});
|
||||
if (!response.ok) return readError(response, "保存模型配置失败");
|
||||
return response.json() as Promise<ManagedModel>;
|
||||
}
|
||||
|
||||
export async function testManagedModel(model: ModelSavePayload): Promise<ModelConnectionTestResult> {
|
||||
const response = await apiFetch(`${getBackendBaseURL()}/api/admin/models/test`, {
|
||||
method: "POST",
|
||||
headers: { "Content-Type": "application/json" },
|
||||
body: JSON.stringify({ model }),
|
||||
});
|
||||
if (!response.ok) return readError(response, "模型连接测试失败");
|
||||
return response.json() as Promise<ModelConnectionTestResult>;
|
||||
}
|
||||
@@ -0,0 +1,38 @@
|
||||
import { useMutation, useQuery, useQueryClient } from "@tanstack/react-query";
|
||||
|
||||
import {
|
||||
createManagedModel,
|
||||
loadManagedModels,
|
||||
testManagedModel,
|
||||
updateManagedModel,
|
||||
type ModelSavePayload,
|
||||
} from "./api";
|
||||
|
||||
export const MANAGED_MODELS_QUERY_KEY = ["admin", "model-config"] as const;
|
||||
|
||||
export function useManagedModels() {
|
||||
return useQuery({
|
||||
queryKey: MANAGED_MODELS_QUERY_KEY,
|
||||
queryFn: loadManagedModels,
|
||||
staleTime: 10_000,
|
||||
refetchOnWindowFocus: false,
|
||||
});
|
||||
}
|
||||
|
||||
export function useSaveManagedModel() {
|
||||
const queryClient = useQueryClient();
|
||||
return useMutation({
|
||||
mutationFn: ({ name, payload, isNew }: { name: string; payload: ModelSavePayload; isNew: boolean }) =>
|
||||
isNew ? createManagedModel(payload) : updateManagedModel(name, payload),
|
||||
onSuccess: async () => {
|
||||
await Promise.all([
|
||||
queryClient.invalidateQueries({ queryKey: MANAGED_MODELS_QUERY_KEY }),
|
||||
queryClient.invalidateQueries({ queryKey: ["models"] }),
|
||||
]);
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
export function useTestManagedModel() {
|
||||
return useMutation({ mutationFn: testManagedModel });
|
||||
}
|
||||
@@ -0,0 +1,2 @@
|
||||
export * from "./api";
|
||||
export * from "./hooks";
|
||||
@@ -0,0 +1,21 @@
|
||||
import type { Model } from "./types";
|
||||
|
||||
export const SINGLE_ROUND_RAG_MODELS: Model[] = [
|
||||
{ id: "single-round-rag-openai3", name: "OpenAI3", model: "OpenAI3", display_name: "Qwen3-32B", supports_thinking: true, execution_profile: "single_round_rag" },
|
||||
{ id: "single-round-rag-openai4", name: "OpenAI4", model: "OpenAI4", display_name: "Qwen3-235B", supports_thinking: true, execution_profile: "single_round_rag" },
|
||||
{ id: "single-round-rag-openai5", name: "OpenAI5", model: "OpenAI5", display_name: "GLM-5", supports_thinking: true, execution_profile: "single_round_rag" },
|
||||
{ id: "single-round-rag-openai6", name: "OpenAI6", model: "OpenAI6", display_name: "Qwen3.5-27B", supports_thinking: true, execution_profile: "single_round_rag" },
|
||||
{ id: "single-round-rag-openai7", name: "OpenAI7", model: "OpenAI7", display_name: "Qwen3-8B", supports_thinking: true, execution_profile: "single_round_rag" },
|
||||
{ id: "single-round-rag-openai9", name: "OpenAI9", model: "OpenAI9", display_name: "四季7B", supports_thinking: true, execution_profile: "single_round_rag" },
|
||||
];
|
||||
|
||||
export function isSingleRoundRagProfile(
|
||||
context: unknown,
|
||||
): boolean {
|
||||
return (
|
||||
typeof context === "object" &&
|
||||
context !== null &&
|
||||
(context as { execution_profile?: unknown }).execution_profile ===
|
||||
"single_round_rag"
|
||||
);
|
||||
}
|
||||
@@ -4,9 +4,11 @@ export interface Model {
|
||||
model: string;
|
||||
display_name: string;
|
||||
description?: string | null;
|
||||
supports_thinking?: boolean;
|
||||
supports_reasoning_effort?: boolean;
|
||||
}
|
||||
supports_thinking?: boolean;
|
||||
supports_reasoning_effort?: boolean;
|
||||
/** Execution profiles handled by the frontend instead of DeerFlow. */
|
||||
execution_profile?: "single_round_rag";
|
||||
}
|
||||
|
||||
export interface TokenUsageSettings {
|
||||
enabled: boolean;
|
||||
|
||||
@@ -23,9 +23,11 @@ export interface AgentThreadContext extends Record<string, unknown> {
|
||||
rag_mode_enabled?: boolean;
|
||||
/** Per-request switch for context compression (上下文压缩). Default off. */
|
||||
summarization_enabled?: boolean;
|
||||
/** Per-conversation switch for knowledge-base RAG injection (phase 2). */
|
||||
knowledge_rag_enabled?: boolean;
|
||||
agent_id?: string;
|
||||
/** Per-conversation switch for knowledge-base RAG injection (phase 2). */
|
||||
knowledge_rag_enabled?: boolean;
|
||||
/** Direct knowledge-base RAG mode; this request never enters Lead Agent. */
|
||||
execution_profile?: "single_round_rag";
|
||||
agent_id?: string;
|
||||
/** @deprecated kept for reading legacy thread metadata; prefer agent_id */
|
||||
agent_name?: string;
|
||||
}
|
||||
|
||||
@@ -0,0 +1,378 @@
|
||||
import type { Message } from "@langchain/langgraph-sdk";
|
||||
import { useCallback, useEffect, useRef, useState } from "react";
|
||||
|
||||
import { readRuntimeEnv } from "@/env";
|
||||
|
||||
const STORAGE_PREFIX = "workspace.single-round-rag.";
|
||||
|
||||
type KnowledgeBaseChunk = {
|
||||
answer?: unknown;
|
||||
docs?: unknown;
|
||||
};
|
||||
|
||||
type RagSource = {
|
||||
id?: string;
|
||||
title: string;
|
||||
type: string;
|
||||
snippet: string;
|
||||
content: string;
|
||||
source?: string;
|
||||
time?: string;
|
||||
url?: string;
|
||||
raw: Record<string, unknown>;
|
||||
};
|
||||
|
||||
/** Same JSON-object framing used by the established GeneralQA single-RAG flow. */
|
||||
function extractJsonObjects(raw: string): { objects: string[]; rest: string } {
|
||||
const objects: string[] = [];
|
||||
let depth = 0;
|
||||
let stringQuote: '"' | "'" | null = null;
|
||||
let escaped = false;
|
||||
let start = -1;
|
||||
for (let index = 0; index < raw.length; index += 1) {
|
||||
const char = raw[index]!;
|
||||
if (stringQuote) {
|
||||
if (escaped) escaped = false;
|
||||
else if (char === "\\") escaped = true;
|
||||
else if (char === stringQuote) stringQuote = null;
|
||||
continue;
|
||||
}
|
||||
if (char === '"' || char === "'") {
|
||||
stringQuote = char;
|
||||
continue;
|
||||
}
|
||||
if (char === "{") {
|
||||
if (depth === 0) start = index;
|
||||
depth += 1;
|
||||
} else if (char === "}") {
|
||||
depth -= 1;
|
||||
if (depth === 0 && start !== -1) {
|
||||
objects.push(raw.slice(start, index + 1));
|
||||
start = -1;
|
||||
}
|
||||
}
|
||||
}
|
||||
return depth > 0 && start !== -1
|
||||
? { objects, rest: raw.slice(start) }
|
||||
: { objects, rest: "" };
|
||||
}
|
||||
|
||||
/** Same `<think>` split semantics as GeneralQA, kept local to avoid loading its demo data. */
|
||||
function parseSingleRagStream(chunks: { answer: string }[]) {
|
||||
let thinkingMarkdown = "";
|
||||
let answerMarkdown = "";
|
||||
let inThink = false;
|
||||
let thinkingComplete = false;
|
||||
for (const { answer } of chunks) {
|
||||
let remaining = answer;
|
||||
while (remaining) {
|
||||
if (inThink) {
|
||||
const endIndex = remaining.indexOf("</think>");
|
||||
if (endIndex === -1) {
|
||||
thinkingMarkdown += remaining;
|
||||
remaining = "";
|
||||
} else {
|
||||
thinkingMarkdown += remaining.slice(0, endIndex);
|
||||
remaining = remaining.slice(endIndex + "</think>".length);
|
||||
inThink = false;
|
||||
thinkingComplete = true;
|
||||
}
|
||||
} else {
|
||||
const startIndex = remaining.indexOf("<think>");
|
||||
if (startIndex === -1) {
|
||||
answerMarkdown += remaining;
|
||||
remaining = "";
|
||||
} else {
|
||||
answerMarkdown += remaining.slice(0, startIndex);
|
||||
remaining = remaining.slice(startIndex + "<think>".length);
|
||||
inThink = true;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
return { thinkingMarkdown, answerMarkdown, thinkingComplete };
|
||||
}
|
||||
|
||||
function storageKey(threadId: string) {
|
||||
return `${STORAGE_PREFIX}${threadId}`;
|
||||
}
|
||||
|
||||
function readStoredMessages(threadId: string): Message[] {
|
||||
if (typeof window === "undefined" || !threadId) return [];
|
||||
try {
|
||||
const raw = window.sessionStorage.getItem(storageKey(threadId));
|
||||
const parsed: unknown = raw ? JSON.parse(raw) : [];
|
||||
return Array.isArray(parsed) ? (parsed as Message[]) : [];
|
||||
} catch {
|
||||
return [];
|
||||
}
|
||||
}
|
||||
|
||||
function writeStoredMessages(threadId: string, messages: Message[]) {
|
||||
if (typeof window === "undefined" || !threadId) return;
|
||||
try {
|
||||
window.sessionStorage.setItem(storageKey(threadId), JSON.stringify(messages));
|
||||
} catch {
|
||||
// Quota errors must never interrupt an already-running answer.
|
||||
}
|
||||
}
|
||||
|
||||
function parseLooseStringValue(rawValue: string): string {
|
||||
try {
|
||||
return JSON.parse(rawValue) as string;
|
||||
} catch {
|
||||
const value = rawValue.trim();
|
||||
if (value.startsWith("'") && value.endsWith("'")) {
|
||||
return value
|
||||
.slice(1, -1)
|
||||
.replace(/\\n/g, "\n")
|
||||
.replace(/\\r/g, "\r")
|
||||
.replace(/\\t/g, "\t")
|
||||
.replace(/\\'/g, "'")
|
||||
.replace(/\\\\/g, "\\");
|
||||
}
|
||||
throw new Error(`无法解析知识库流片段：${rawValue}`);
|
||||
}
|
||||
}
|
||||
|
||||
function parseKnowledgeBaseChunk(jsonText: string): KnowledgeBaseChunk {
|
||||
try {
|
||||
return JSON.parse(jsonText) as KnowledgeBaseChunk;
|
||||
} catch {
|
||||
const objectText = jsonText.trim();
|
||||
const answerMatch = objectText.match(/^\{\s*answer\s*:\s*([\s\S]*)\s*\}$/);
|
||||
if (answerMatch) {
|
||||
return { answer: parseLooseStringValue(answerMatch[1]!.trim()) };
|
||||
}
|
||||
const docsMatch = objectText.match(/^\{\s*docs\s*:\s*([\s\S]*)\s*\}$/);
|
||||
if (docsMatch) {
|
||||
const rawValue = docsMatch[1]!.trim();
|
||||
try {
|
||||
return { docs: JSON.parse(rawValue) };
|
||||
} catch {
|
||||
return { docs: parseLooseStringValue(rawValue) };
|
||||
}
|
||||
}
|
||||
const normalizedText = objectText.replace(
|
||||
/([{,]\s*)(answer|docs)(\s*:)/g,
|
||||
'$1"$2"$3',
|
||||
);
|
||||
return JSON.parse(normalizedText) as KnowledgeBaseChunk;
|
||||
}
|
||||
}
|
||||
|
||||
function flattenDocs(value: unknown): Record<string, unknown>[] {
|
||||
if (Array.isArray(value)) {
|
||||
return value.flatMap((item) => flattenDocs(item));
|
||||
}
|
||||
if (!value || typeof value !== "object") return [];
|
||||
const record = value as Record<string, unknown>;
|
||||
if ("docs" in record) return flattenDocs(record.docs);
|
||||
return [record];
|
||||
}
|
||||
|
||||
function normalizeSources(value: unknown): RagSource[] {
|
||||
return flattenDocs(value).map((item, index) => {
|
||||
const title = String(item.m_title ?? item.title ?? item.source ?? `参考资料 ${index + 1}`);
|
||||
const content = String(item.m_content ?? item.content ?? item.snippet ?? "");
|
||||
const source = item.source1 ?? item.source;
|
||||
const rawUrl = item.url ?? item.link ?? item.href;
|
||||
return {
|
||||
id: String(item.recUuid ?? item.rec_uuid ?? item.uuid ?? item.id ?? "") || undefined,
|
||||
title,
|
||||
type: String(item.source1 ?? item.type ?? "知识库"),
|
||||
snippet: content,
|
||||
content,
|
||||
source: typeof source === "string" ? source : undefined,
|
||||
time: typeof item.m_publish === "string" ? item.m_publish : undefined,
|
||||
url: typeof rawUrl === "string" && /^https?:\/\//i.test(rawUrl) ? rawUrl : undefined,
|
||||
raw: item,
|
||||
};
|
||||
});
|
||||
}
|
||||
|
||||
function createId(prefix: string) {
|
||||
return `${prefix}-${crypto.randomUUID()}`;
|
||||
}
|
||||
|
||||
function isAbortError(error: unknown) {
|
||||
return error instanceof DOMException && error.name === "AbortError";
|
||||
}
|
||||
|
||||
export function isSingleRoundRagMessage(message: Message): boolean {
|
||||
return message.additional_kwargs?.single_round_rag === true;
|
||||
}
|
||||
|
||||
/**
|
||||
* Frontend-only adapter for the existing general-QA single-round endpoint.
|
||||
* It deliberately never submits a run to DeerFlow or mutates the Lead Agent.
|
||||
*/
|
||||
export function useSingleRoundRagMessages(threadId: string) {
|
||||
const [messages, setMessages] = useState<Message[]>(() => readStoredMessages(threadId));
|
||||
const [isLoading, setIsLoading] = useState(false);
|
||||
const [error, setError] = useState<string | null>(null);
|
||||
const controllerRef = useRef<AbortController | null>(null);
|
||||
|
||||
useEffect(() => {
|
||||
setMessages(readStoredMessages(threadId));
|
||||
setError(null);
|
||||
}, [threadId]);
|
||||
|
||||
useEffect(() => () => controllerRef.current?.abort(), []);
|
||||
|
||||
const updateMessages = useCallback(
|
||||
(targetThreadId: string, updater: (current: Message[]) => Message[]) => {
|
||||
setMessages((current) => {
|
||||
const next = updater(current);
|
||||
writeStoredMessages(targetThreadId, next);
|
||||
return next;
|
||||
});
|
||||
},
|
||||
[],
|
||||
);
|
||||
|
||||
const stop = useCallback(() => controllerRef.current?.abort(), []);
|
||||
|
||||
const submit = useCallback(
|
||||
async ({
|
||||
targetThreadId,
|
||||
query,
|
||||
modelName,
|
||||
}: {
|
||||
targetThreadId: string;
|
||||
query: string;
|
||||
modelName: string;
|
||||
}) => {
|
||||
if (isLoading) return;
|
||||
const question = query.trim();
|
||||
if (!question) return;
|
||||
|
||||
const humanId = createId("single-rag-human");
|
||||
const assistantId = createId("single-rag-ai");
|
||||
const sourceId = createId("single-rag-sources");
|
||||
const clientTs = new Date().toISOString();
|
||||
const userMessage: Message = {
|
||||
id: humanId,
|
||||
type: "human",
|
||||
content: [{ type: "text", text: question }],
|
||||
additional_kwargs: { client_ts: clientTs, single_round_rag: true },
|
||||
};
|
||||
const initialAssistant: Message = {
|
||||
id: assistantId,
|
||||
type: "ai",
|
||||
content: "",
|
||||
additional_kwargs: { single_round_rag: true },
|
||||
};
|
||||
updateMessages(targetThreadId, (current) => [...current, userMessage, initialAssistant]);
|
||||
setError(null);
|
||||
setIsLoading(true);
|
||||
|
||||
const controller = new AbortController();
|
||||
controllerRef.current = controller;
|
||||
const answerChunks: { answer: string }[] = [];
|
||||
let sources: RagSource[] = [];
|
||||
|
||||
const refreshAssistant = () => {
|
||||
const trace = parseSingleRagStream(answerChunks);
|
||||
updateMessages(targetThreadId, (current) => {
|
||||
const next = current.filter((message) => message.id !== sourceId);
|
||||
const assistantIndex = next.findIndex((message) => message.id === assistantId);
|
||||
if (assistantIndex < 0) return current;
|
||||
next[assistantIndex] = {
|
||||
...next[assistantIndex],
|
||||
content: trace.answerMarkdown,
|
||||
additional_kwargs: {
|
||||
...next[assistantIndex]!.additional_kwargs,
|
||||
single_round_rag: true,
|
||||
reasoning_content: trace.thinkingMarkdown || null,
|
||||
},
|
||||
} as Message;
|
||||
if (sources.length > 0) {
|
||||
next.splice(assistantIndex, 0, {
|
||||
id: sourceId,
|
||||
type: "tool",
|
||||
name: "knowledge_base_search",
|
||||
content: { results: sources },
|
||||
additional_kwargs: {
|
||||
hide_from_ui: true,
|
||||
single_round_rag: true,
|
||||
},
|
||||
} as unknown as Message);
|
||||
}
|
||||
return next;
|
||||
});
|
||||
};
|
||||
|
||||
try {
|
||||
const chatUrl = readRuntimeEnv("VITE_AIPI_CHAT_URL") || "";
|
||||
if (!chatUrl) {
|
||||
throw new Error("未配置 VITE_AIPI_CHAT_URL，无法发起单轮 RAG 问答。");
|
||||
}
|
||||
const response = await fetch(
|
||||
`${chatUrl.replace(/\/$/, "")}/chat/knowledge_base_chat_tiaoshi`,
|
||||
{
|
||||
method: "POST",
|
||||
headers: { "Content-Type": "application/json" },
|
||||
signal: controller.signal,
|
||||
body: JSON.stringify({
|
||||
query: question,
|
||||
knowledge_base_name: [],
|
||||
top_k: 5,
|
||||
score_threshold: 0.5,
|
||||
history: [],
|
||||
stream: true,
|
||||
model_name: modelName,
|
||||
temperature: 0.3,
|
||||
prompt_name: "default",
|
||||
local_doc_url: false,
|
||||
}),
|
||||
},
|
||||
);
|
||||
if (!response.ok) throw new Error(`知识库服务请求失败（${response.status}）`);
|
||||
if (!response.body) throw new Error("知识库服务未返回流式响应。");
|
||||
|
||||
const reader = response.body.getReader();
|
||||
const decoder = new TextDecoder();
|
||||
let buffer = "";
|
||||
while (true) {
|
||||
const { done, value } = await reader.read();
|
||||
if (done) break;
|
||||
buffer += decoder.decode(value, { stream: true });
|
||||
const extracted = extractJsonObjects(buffer);
|
||||
buffer = extracted.rest;
|
||||
for (const jsonText of extracted.objects) {
|
||||
const chunk = parseKnowledgeBaseChunk(jsonText);
|
||||
if (typeof chunk.answer === "string") answerChunks.push({ answer: chunk.answer });
|
||||
if (chunk.docs !== undefined) sources = normalizeSources(chunk.docs);
|
||||
refreshAssistant();
|
||||
}
|
||||
}
|
||||
} catch (caught) {
|
||||
if (!isAbortError(caught)) {
|
||||
const message = caught instanceof Error ? caught.message : "单轮 RAG 问答失败。";
|
||||
setError(message);
|
||||
updateMessages(targetThreadId, (current) =>
|
||||
current.map((item) =>
|
||||
item.id === assistantId
|
||||
? ({
|
||||
...item,
|
||||
content: answerChunks.length > 0 ? item.content : message,
|
||||
additional_kwargs: {
|
||||
...item.additional_kwargs,
|
||||
single_round_rag: true,
|
||||
},
|
||||
} as Message)
|
||||
: item,
|
||||
),
|
||||
);
|
||||
}
|
||||
} finally {
|
||||
if (controllerRef.current === controller) controllerRef.current = null;
|
||||
setIsLoading(false);
|
||||
}
|
||||
},
|
||||
[isLoading, updateMessages],
|
||||
);
|
||||
|
||||
return { messages, isLoading, error, submit, stop };
|
||||
}
|
||||
@@ -44,6 +44,10 @@ import { PanelLeftOpen } from "lucide-react";
|
||||
import { useAgentSelector, useAgentsPaginated } from "@/core/agents";
|
||||
import { useI18n } from "@/core/i18n/hooks";
|
||||
import { useModels } from "@/core/models/hooks";
|
||||
import {
|
||||
isSingleRoundRagProfile,
|
||||
SINGLE_ROUND_RAG_MODELS,
|
||||
} from "@/core/models/single-round-rag";
|
||||
import { useNotification } from "@/core/notification/hooks";
|
||||
import {
|
||||
getLastAgentId,
|
||||
@@ -52,6 +56,7 @@ import {
|
||||
useThreadSettings,
|
||||
} from "@/core/settings";
|
||||
import { useThreadStream } from "@/core/threads/hooks";
|
||||
import { useSingleRoundRagMessages } from "@/core/threads/use-single-round-rag";
|
||||
import { useLayoutMode } from "@/core/system-settings/hooks";
|
||||
import { useRouteBase } from "@/core/route-base";
|
||||
import { textOfMessage } from "@/core/threads/utils";
|
||||
@@ -226,6 +231,20 @@ export default function ChatPage() {
|
||||
}
|
||||
},
|
||||
});
|
||||
const singleRoundRag = useSingleRoundRagMessages(threadId);
|
||||
const singleRoundRagSelected = isSingleRoundRagProfile(settings.context);
|
||||
const displayThread = useMemo(() => {
|
||||
if (singleRoundRag.messages.length === 0 && !singleRoundRag.isLoading) {
|
||||
return thread;
|
||||
}
|
||||
const messages = [...thread.messages, ...singleRoundRag.messages];
|
||||
return {
|
||||
...thread,
|
||||
values: { ...thread.values, messages },
|
||||
messages,
|
||||
isLoading: thread.isLoading || singleRoundRag.isLoading,
|
||||
} as typeof thread;
|
||||
}, [singleRoundRag.isLoading, singleRoundRag.messages, thread]);
|
||||
|
||||
const writingArtifactPath = useMemo(() => {
|
||||
const isMarkdown = (path: string | null | undefined) =>
|
||||
@@ -240,6 +259,46 @@ export default function ChatPage() {
|
||||
|
||||
const handleSubmit = useCallback(
|
||||
(message: PromptInputMessage) => {
|
||||
if (singleRoundRagSelected) {
|
||||
if (message.files && message.files.length > 0) {
|
||||
toast.error("单轮 RAG 问答暂不支持附件，请移除附件后重试。");
|
||||
return;
|
||||
}
|
||||
setHasStartedConversation(true);
|
||||
window.dispatchEvent(
|
||||
new CustomEvent("workspace:sandbox-resume-autoscroll"),
|
||||
);
|
||||
void (async () => {
|
||||
try {
|
||||
let targetThreadId = threadId;
|
||||
if (isNewThread) {
|
||||
const created = await getAPIClient(isMock).threads.create({});
|
||||
targetThreadId = created.thread_id;
|
||||
setThreadId(targetThreadId);
|
||||
setIsNewThread(false);
|
||||
history.replaceState(
|
||||
null,
|
||||
"",
|
||||
`${window.location.pathname}${window.location.search}#${routeBase}/chats/${targetThreadId}`,
|
||||
);
|
||||
const title = message.text.trim().replace(/\s+/g, " ").slice(0, 60);
|
||||
if (title) {
|
||||
void getAPIClient(isMock).threads
|
||||
.updateState(targetThreadId, { values: { title } })
|
||||
.catch(() => undefined);
|
||||
}
|
||||
}
|
||||
await singleRoundRag.submit({
|
||||
targetThreadId,
|
||||
query: message.text,
|
||||
modelName: settings.context.model_name ?? "OpenAI3",
|
||||
});
|
||||
} catch (error) {
|
||||
toast.error(error instanceof Error ? error.message : "单轮 RAG 问答失败。");
|
||||
}
|
||||
})();
|
||||
return;
|
||||
}
|
||||
// If the user picked an agent before sending the opening turn, hand off
|
||||
// to AgentChatPage so the conversation runs against that agent. We stash
|
||||
// the pending text in sessionStorage because the next/navigation shim
|
||||
@@ -332,7 +391,13 @@ export default function ChatPage() {
|
||||
isMock,
|
||||
router,
|
||||
settings.context.writing_mode,
|
||||
settings.context.model_name,
|
||||
writingArtifactPath,
|
||||
singleRoundRag,
|
||||
singleRoundRagSelected,
|
||||
routeBase,
|
||||
setIsNewThread,
|
||||
setThreadId,
|
||||
],
|
||||
);
|
||||
|
||||
@@ -421,8 +486,12 @@ export default function ChatPage() {
|
||||
saveLastAgentId(agentId);
|
||||
}, []);
|
||||
const handleStop = useCallback(async () => {
|
||||
if (singleRoundRag.isLoading) {
|
||||
singleRoundRag.stop();
|
||||
return;
|
||||
}
|
||||
await thread.stop();
|
||||
}, [thread]);
|
||||
}, [singleRoundRag, thread]);
|
||||
|
||||
const promptPrefixProp = useMemo(() => {
|
||||
if (!promptPrefixSettings?.enabled) return undefined;
|
||||
@@ -468,7 +537,7 @@ export default function ChatPage() {
|
||||
const compactActionsHidden = isHideChromeActive();
|
||||
|
||||
return (
|
||||
<ThreadContext.Provider value={{ thread, isMock }}>
|
||||
<ThreadContext.Provider value={{ thread: displayThread, isMock }}>
|
||||
<ChatBox threadId={threadId}>
|
||||
<div
|
||||
className={cn(
|
||||
@@ -486,7 +555,7 @@ export default function ChatPage() {
|
||||
)}
|
||||
>
|
||||
<div className="flex w-full items-center text-sm font-medium">
|
||||
<ThreadTitle threadId={threadId} thread={thread} />
|
||||
<ThreadTitle threadId={threadId} thread={displayThread} />
|
||||
</div>
|
||||
<div className="flex items-center gap-2">
|
||||
{/* 仅在收起态显示「展开历史」钮；展开后由侧栏内（新对话右侧）的收起钮负责收起。 */}
|
||||
@@ -505,7 +574,7 @@ export default function ChatPage() {
|
||||
)}
|
||||
<TokenUsageIndicator
|
||||
enabled={tokenUsageEnabled}
|
||||
messages={thread.messages}
|
||||
messages={displayThread.messages}
|
||||
/>
|
||||
<ExportTrigger threadId={threadId} />
|
||||
<KnowledgeSaveTrigger threadId={threadId} />
|
||||
@@ -543,7 +612,7 @@ export default function ChatPage() {
|
||||
<MessageList
|
||||
className={cn("size-full", "pt-10")}
|
||||
threadId={threadId}
|
||||
thread={thread}
|
||||
thread={displayThread}
|
||||
paddingBottom={messageListPaddingBottom}
|
||||
tokenUsageEnabled={tokenUsageEnabled}
|
||||
// 输入框为底部绝对浮层（约 160px 高），按钮上移到其上方居中。
|
||||
@@ -585,9 +654,9 @@ export default function ChatPage() {
|
||||
<div className="absolute right-0 bottom-0 left-0">
|
||||
<TodoList
|
||||
className="bg-background/5"
|
||||
todos={thread.values.todos ?? []}
|
||||
todos={displayThread.values.todos ?? []}
|
||||
hidden={
|
||||
!thread.values.todos || thread.values.todos.length === 0
|
||||
!displayThread.values.todos || displayThread.values.todos.length === 0
|
||||
}
|
||||
/>
|
||||
</div>
|
||||
@@ -612,15 +681,17 @@ export default function ChatPage() {
|
||||
status={
|
||||
thread.error
|
||||
? "error"
|
||||
: thread.isLoading
|
||||
: displayThread.isLoading
|
||||
? "streaming"
|
||||
: "ready"
|
||||
}
|
||||
context={settings.context}
|
||||
extraModels={SINGLE_ROUND_RAG_MODELS}
|
||||
disabled={
|
||||
env.NEXT_PUBLIC_STATIC_WEBSITE_ONLY === "true" ||
|
||||
isUploading ||
|
||||
agentHandoffPending
|
||||
agentHandoffPending ||
|
||||
singleRoundRag.isLoading
|
||||
}
|
||||
onContextChange={(context) =>
|
||||
setSettings("context", context)
|
||||
@@ -664,7 +735,7 @@ export default function ChatPage() {
|
||||
*/}
|
||||
{!isLandingState && (
|
||||
<QuestionNavigator
|
||||
messages={thread.messages}
|
||||
messages={displayThread.messages}
|
||||
className="top-16 right-0 bottom-48 left-0"
|
||||
/>
|
||||
)}
|
||||
|
||||
@@ -0,0 +1,246 @@
|
||||
import {
|
||||
BotIcon,
|
||||
ChevronDownIcon,
|
||||
CircleAlertIcon,
|
||||
CpuIcon,
|
||||
EyeIcon,
|
||||
KeyRoundIcon,
|
||||
LoaderCircleIcon,
|
||||
PlusIcon,
|
||||
RefreshCwIcon,
|
||||
Settings2Icon,
|
||||
SparklesIcon,
|
||||
WrenchIcon,
|
||||
} from "lucide-react";
|
||||
import { useMemo, useState, type ReactNode } from "react";
|
||||
import {
|
||||
Button as TButton,
|
||||
Dialog,
|
||||
Input,
|
||||
InputNumber,
|
||||
MessagePlugin,
|
||||
Select,
|
||||
Switch,
|
||||
Tag,
|
||||
Textarea,
|
||||
} from "tdesign-react";
|
||||
|
||||
import {
|
||||
type ManagedModel,
|
||||
type ModelSavePayload,
|
||||
useManagedModels,
|
||||
useSaveManagedModel,
|
||||
useTestManagedModel,
|
||||
} from "@/core/model-management";
|
||||
import {
|
||||
WorkspaceBody,
|
||||
WorkspaceContainer,
|
||||
WorkspaceHeader,
|
||||
} from "@/components/workspace/workspace-container";
|
||||
|
||||
type ModelDraft = ModelSavePayload & {
|
||||
api_key: string;
|
||||
keep_existing_key: boolean;
|
||||
extra_settings_text: string;
|
||||
};
|
||||
|
||||
const PROVIDER_OPTIONS = [
|
||||
{ label: "OpenAI 兼容接口", value: "langchain_openai:ChatOpenAI" },
|
||||
{ label: "Anthropic", value: "langchain_anthropic:ChatAnthropic" },
|
||||
{ label: "Google Gemini", value: "langchain_google_genai:ChatGoogleGenerativeAI" },
|
||||
{ label: "Groq", value: "langchain_groq:ChatGroq" },
|
||||
{ label: "Ollama", value: "langchain_ollama:ChatOllama" },
|
||||
];
|
||||
|
||||
const emptyDraft = (): ModelDraft => ({
|
||||
name: "",
|
||||
display_name: "",
|
||||
description: "",
|
||||
use: "langchain_openai:ChatOpenAI",
|
||||
model: "",
|
||||
api_key: "",
|
||||
keep_existing_key: false,
|
||||
clear_api_key: false,
|
||||
base_url: "",
|
||||
request_timeout: 30,
|
||||
max_retries: 2,
|
||||
max_tokens: undefined,
|
||||
temperature: undefined,
|
||||
stream_usage: true,
|
||||
supports_thinking: false,
|
||||
supports_reasoning_effort: false,
|
||||
supports_vision: false,
|
||||
when_thinking_enabled: null,
|
||||
when_thinking_disabled: null,
|
||||
thinking: null,
|
||||
use_responses_api: null,
|
||||
output_version: null,
|
||||
extra_settings: {},
|
||||
extra_settings_text: "{}",
|
||||
});
|
||||
|
||||
function draftFromModel(model: ManagedModel): ModelDraft {
|
||||
return {
|
||||
...model,
|
||||
api_key: "",
|
||||
keep_existing_key: model.api_key_configured,
|
||||
clear_api_key: false,
|
||||
extra_settings_text: JSON.stringify(model.extra_settings ?? {}, null, 2),
|
||||
};
|
||||
}
|
||||
|
||||
function optionalString(value: string | null | undefined): string | null {
|
||||
const normalized = (value ?? "").trim();
|
||||
return normalized || null;
|
||||
}
|
||||
|
||||
function toNumber(value: number | string | null | undefined): number | undefined {
|
||||
if (value === "" || value === null || value === undefined) return undefined;
|
||||
const parsed = Number(value);
|
||||
return Number.isFinite(parsed) ? parsed : undefined;
|
||||
}
|
||||
|
||||
function buildPayload(draft: ModelDraft): ModelSavePayload {
|
||||
let extraSettings: Record<string, unknown> = {};
|
||||
try {
|
||||
const parsed = JSON.parse(draft.extra_settings_text || "{}") as unknown;
|
||||
if (!parsed || Array.isArray(parsed) || typeof parsed !== "object") {
|
||||
throw new Error("额外参数必须是 JSON 对象");
|
||||
}
|
||||
extraSettings = parsed as Record<string, unknown>;
|
||||
} catch (error) {
|
||||
throw error instanceof Error ? error : new Error("额外参数不是有效 JSON");
|
||||
}
|
||||
const payload: ModelSavePayload = {
|
||||
name: draft.name.trim(),
|
||||
display_name: optionalString(draft.display_name),
|
||||
description: optionalString(draft.description),
|
||||
use: draft.use.trim(),
|
||||
model: draft.model.trim(),
|
||||
clear_api_key: Boolean(draft.clear_api_key),
|
||||
base_url: optionalString(draft.base_url),
|
||||
request_timeout: toNumber(draft.request_timeout),
|
||||
max_retries: toNumber(draft.max_retries),
|
||||
max_tokens: toNumber(draft.max_tokens),
|
||||
temperature: toNumber(draft.temperature),
|
||||
stream_usage: draft.stream_usage,
|
||||
supports_thinking: draft.supports_thinking,
|
||||
supports_reasoning_effort: draft.supports_reasoning_effort,
|
||||
supports_vision: draft.supports_vision,
|
||||
when_thinking_enabled: draft.when_thinking_enabled,
|
||||
when_thinking_disabled: draft.when_thinking_disabled,
|
||||
thinking: draft.thinking,
|
||||
use_responses_api: draft.use_responses_api,
|
||||
output_version: optionalString(draft.output_version),
|
||||
extra_settings: extraSettings,
|
||||
};
|
||||
if (draft.api_key.trim()) payload.api_key = draft.api_key.trim();
|
||||
return payload;
|
||||
}
|
||||
|
||||
function StatCard({ icon: Icon, label, value }: { icon: typeof CpuIcon; label: string; value: number }) {
|
||||
return (
|
||||
<div className="flex items-center gap-3 rounded-lg border border-border bg-card px-4 py-3 shadow-sm">
|
||||
<span className="flex size-10 items-center justify-center rounded-lg bg-primary/10 text-primary"><Icon className="size-5" /></span>
|
||||
<div><div className="text-xl font-semibold tabular-nums">{value}</div><div className="text-xs text-muted-foreground">{label}</div></div>
|
||||
</div>
|
||||
);
|
||||
}
|
||||
|
||||
function Field({ label, required = false, children }: { label: string; required?: boolean; children: ReactNode }) {
|
||||
return <label className="space-y-1.5 text-sm font-medium">{label}{required && <span className="ml-1 text-destructive">*</span>}{children}</label>;
|
||||
}
|
||||
|
||||
function ModelEditor({ model, onClose }: { model: ManagedModel | null; onClose: () => void }) {
|
||||
const isNew = model === null;
|
||||
const [draft, setDraft] = useState<ModelDraft>(() => model ? draftFromModel(model) : emptyDraft());
|
||||
const [advanced, setAdvanced] = useState(false);
|
||||
const [testResult, setTestResult] = useState<{ success: boolean; latency: number; message: string; preview?: string | null } | null>(null);
|
||||
const save = useSaveManagedModel();
|
||||
const test = useTestManagedModel();
|
||||
const patch = (next: Partial<ModelDraft>) => setDraft((current) => ({ ...current, ...next }));
|
||||
|
||||
const handleTest = async () => {
|
||||
try {
|
||||
const payload = buildPayload(draft);
|
||||
const result = await test.mutateAsync(payload);
|
||||
setTestResult({ success: result.success, latency: result.latency_ms, message: result.message, preview: result.preview });
|
||||
if (!result.success) return void MessagePlugin.error(`连接失败：${result.message}`);
|
||||
MessagePlugin.success(`连接成功${result.latency_ms ? `，${result.latency_ms} ms` : ""}`);
|
||||
} catch (error) {
|
||||
MessagePlugin.error(error instanceof Error ? error.message : "连接测试失败");
|
||||
}
|
||||
};
|
||||
|
||||
const handleSave = async () => {
|
||||
try {
|
||||
const payload = buildPayload(draft);
|
||||
if (!payload.name || !payload.use || !payload.model) {
|
||||
return void MessagePlugin.warning("请填写模型标识、驱动与实际模型名");
|
||||
}
|
||||
await save.mutateAsync({ name: model?.name ?? payload.name, payload, isNew });
|
||||
MessagePlugin.success(isNew ? "模型已添加，所有页面的模型选择器已刷新" : "模型配置已保存并立即生效");
|
||||
onClose();
|
||||
} catch (error) {
|
||||
MessagePlugin.error(error instanceof Error ? error.message : "保存模型配置失败");
|
||||
}
|
||||
};
|
||||
|
||||
return (
|
||||
<Dialog
|
||||
visible
|
||||
header={isNew ? "添加模型" : `编辑模型：${model?.display_name || model?.name}`}
|
||||
placement="center"
|
||||
width={860}
|
||||
zIndex={3200}
|
||||
confirmBtn={{ content: isNew ? "添加并启用" : "保存配置", theme: "primary" }}
|
||||
cancelBtn="取消"
|
||||
confirmLoading={save.isPending}
|
||||
onClose={onClose}
|
||||
onConfirm={() => void handleSave()}
|
||||
onCancel={onClose}
|
||||
>
|
||||
<div className="max-h-[70vh] space-y-5 overflow-y-auto px-1 py-1 pr-3">
|
||||
<div className="rounded-lg border border-primary/20 bg-primary/5 px-3 py-2 text-sm text-muted-foreground">
|
||||
保存会直接更新当前部署的 <code className="rounded bg-background px-1 py-0.5">config.yaml</code> 模型段，自动备份并热刷新；不会中断已开始的会话。
|
||||
</div>
|
||||
<section className="grid gap-4 md:grid-cols-2">
|
||||
<Field label="稳定标识" required><Input value={draft.name} disabled={!isNew} placeholder="例如 deepseek-chat" tips={isNew ? "保存后不可修改，用于智能体和运行配置引用。" : "稳定标识不可修改，避免已有配置失效。"} onChange={(value) => patch({ name: String(value) })} /></Field>
|
||||
<Field label="显示名称"><Input value={draft.display_name ?? ""} placeholder="例如 DeepSeek Chat" onChange={(value) => patch({ display_name: String(value) })} /></Field>
|
||||
<Field label="模型驱动" required><Select value={draft.use} options={PROVIDER_OPTIONS} filterable creatable inputProps={{ placeholder: "选择或输入 use 类路径" }} onChange={(value) => patch({ use: String(value) })} /></Field>
|
||||
<Field label="实际模型名" required><Input value={draft.model} placeholder="例如 deepseek-chat" onChange={(value) => patch({ model: String(value) })} /></Field>
|
||||
<Field label="Base URL"><Input value={draft.base_url ?? ""} type="url" placeholder="https://api.example.com/v1" onChange={(value) => patch({ base_url: String(value) })} /></Field>
|
||||
<Field label="API Key"><Input value={draft.api_key} type="password" placeholder={draft.keep_existing_key && !draft.clear_api_key ? "已配置，留空则保持不变" : "可填写密钥或 $ENVIRONMENT_VARIABLE"} onChange={(value) => patch({ api_key: String(value), clear_api_key: false })} /></Field>
|
||||
{draft.keep_existing_key && !draft.api_key && <div className="md:col-span-2 flex items-center gap-2 text-xs text-muted-foreground"><KeyRoundIcon className="size-3.5" />已配置密钥{model?.api_key_source ? `（环境变量 ${model.api_key_source}）` : ""}，不会回显。<TButton theme="default" variant="text" size="small" onClick={() => patch({ clear_api_key: !draft.clear_api_key })}>{draft.clear_api_key ? "撤销清除" : "清除密钥"}</TButton></div>}
|
||||
<div className="md:col-span-2"><Field label="描述"><Textarea value={draft.description ?? ""} autosize={{ minRows: 2, maxRows: 4 }} placeholder="说明模型适用的任务与限制" onChange={(value) => patch({ description: String(value) })} /></Field></div>
|
||||
</section>
|
||||
<section className="rounded-lg border border-border p-4">
|
||||
<div className="mb-2"><h3 className="font-medium">能力声明</h3><p className="mt-1 text-xs text-muted-foreground">这些标识决定哪些页面会向模型提供思考、视觉或工具调用入口。</p></div>
|
||||
<div className="flex flex-wrap gap-x-5 gap-y-3 text-sm"><span className="flex items-center gap-2"><Switch value={draft.supports_thinking} onChange={(value) => patch({ supports_thinking: Boolean(value) })} />支持思考</span><span className="flex items-center gap-2"><Switch value={draft.supports_vision} onChange={(value) => patch({ supports_vision: Boolean(value) })} />支持视觉</span><span className="flex items-center gap-2"><Switch value={draft.supports_reasoning_effort} onChange={(value) => patch({ supports_reasoning_effort: Boolean(value) })} />推理档位</span></div>
|
||||
</section>
|
||||
<div className="flex items-center justify-between border-t pt-4"><TButton theme="default" variant="outline" loading={test.isPending} onClick={() => void handleTest()}><WrenchIcon className="mr-1.5 size-4" />测试连接</TButton><TButton theme="default" variant="text" onClick={() => setAdvanced((value) => !value)}><Settings2Icon className="mr-1.5 size-4" />高级配置 <ChevronDownIcon className={`ml-1 size-4 transition-transform ${advanced ? "rotate-180" : ""}`} /></TButton></div>
|
||||
{testResult && <div className={`rounded-lg border px-3 py-2 text-sm ${testResult.success ? "border-primary/20 bg-primary/5 text-foreground" : "border-destructive/25 bg-destructive/5 text-destructive"}`}><div className="font-medium">{testResult.success ? "连接成功" : "连接失败"}{testResult.success && testResult.latency ? ` · ${testResult.latency} ms` : ""}</div><p className="mt-1 break-words text-xs opacity-80">{testResult.preview || testResult.message}</p></div>}
|
||||
{advanced && <section className="space-y-4 rounded-lg border border-dashed border-border bg-muted/10 p-4"><div className="grid gap-4 md:grid-cols-3"><Field label="请求超时（秒）"><InputNumber value={draft.request_timeout ?? ""} min={1} max={600} onChange={(value) => patch({ request_timeout: toNumber(value) })} /></Field><Field label="最大重试次数"><InputNumber value={draft.max_retries ?? ""} min={0} max={20} onChange={(value) => patch({ max_retries: toNumber(value) })} /></Field><Field label="最大输出 Token"><InputNumber value={draft.max_tokens ?? ""} min={1} max={1_000_000} onChange={(value) => patch({ max_tokens: toNumber(value) })} /></Field><Field label="温度"><InputNumber value={draft.temperature ?? ""} min={0} max={2} step={0.1} decimalPlaces={1} onChange={(value) => patch({ temperature: toNumber(value) })} /></Field><span className="flex items-end gap-2 pb-1 text-sm font-medium"><Switch value={draft.stream_usage ?? true} onChange={(value) => patch({ stream_usage: Boolean(value) })} />流式 Token 统计</span></div><Field label="额外参数 JSON"><Textarea value={draft.extra_settings_text} autosize={{ minRows: 4, maxRows: 10 }} placeholder={'{\n "extra_body": {}\n}'} onChange={(value) => patch({ extra_settings_text: String(value) })} /><p className="text-xs font-normal text-muted-foreground">保留并编辑配置文件中未在基础表单展示的兼容参数。必须为 JSON 对象。</p></Field></section>}
|
||||
</div>
|
||||
</Dialog>
|
||||
);
|
||||
}
|
||||
|
||||
export default function ModelManagementPage() {
|
||||
const { data, isLoading, error, refetch, isFetching } = useManagedModels();
|
||||
const [search, setSearch] = useState("");
|
||||
const [filter, setFilter] = useState<"all" | "thinking" | "vision">("all");
|
||||
const [editor, setEditor] = useState<ManagedModel | null | "new">(null);
|
||||
const models = useMemo(() => {
|
||||
const keyword = search.trim().toLowerCase();
|
||||
return (data?.models ?? []).filter((model) => {
|
||||
const matchesFilter = filter === "all" || (filter === "thinking" ? model.supports_thinking : model.supports_vision);
|
||||
const matchesSearch = !keyword || [model.name, model.display_name, model.model, model.use].filter(Boolean).join(" ").toLowerCase().includes(keyword);
|
||||
return matchesFilter && matchesSearch;
|
||||
});
|
||||
}, [data?.models, filter, search]);
|
||||
const stats = useMemo(() => ({ total: data?.models.length ?? 0, thinking: data?.models.filter((item) => item.supports_thinking).length ?? 0, vision: data?.models.filter((item) => item.supports_vision).length ?? 0 }), [data?.models]);
|
||||
return (
|
||||
<WorkspaceContainer><WorkspaceHeader /><WorkspaceBody><main className="flex size-full min-h-0 flex-col overflow-y-auto bg-background px-6 py-5 md:px-8"><div className="mx-auto flex w-full max-w-7xl flex-col gap-5"><header className="flex flex-wrap items-start justify-between gap-4"><div><div className="flex items-center gap-2"><span className="flex size-9 items-center justify-center rounded-lg bg-primary/10 text-primary"><CpuIcon className="size-5" /></span><h1 className="text-2xl font-semibold tracking-tight">模型配置</h1></div><p className="mt-2 max-w-3xl text-sm leading-6 text-muted-foreground">管理当前部署 <code className="rounded bg-muted px-1 py-0.5">config.yaml</code> 中的对话模型。保存后所有页面立即可选；密钥仅可写入，永不回显。</p></div><div className="flex flex-wrap items-center gap-2"><TButton theme="default" variant="outline" loading={isFetching} onClick={() => void refetch()}><RefreshCwIcon className="mr-1.5 size-4" />刷新</TButton><TButton theme="primary" onClick={() => setEditor("new")}><PlusIcon className="mr-1.5 size-4" />添加模型</TButton></div></header><div className="grid gap-3 sm:grid-cols-3"><StatCard icon={BotIcon} label="已配置模型" value={stats.total} /><StatCard icon={SparklesIcon} label="支持思考" value={stats.thinking} /><StatCard icon={EyeIcon} label="支持视觉" value={stats.vision} /></div><section className="rounded-xl border border-border bg-card shadow-sm"><div className="flex flex-wrap items-center justify-between gap-3 border-b px-4 py-3"><div className="flex flex-wrap items-center gap-2">{([ ["all", "全部"], ["thinking", "支持思考"], ["vision", "支持视觉"] ] as const).map(([value, label]) => <TButton key={value} theme={filter === value ? "primary" : "default"} variant={filter === value ? "base" : "text"} size="small" onClick={() => setFilter(value)}>{label}</TButton>)}</div><Input value={search} placeholder="搜索名称、模型或驱动" clearable className="w-full sm:w-72" onChange={(value) => setSearch(String(value))} /></div>{isLoading && <div className="flex h-52 items-center justify-center gap-2 text-sm text-muted-foreground"><LoaderCircleIcon className="size-4 animate-spin" />加载模型配置…</div>}{error && <div className="m-4 flex items-start gap-2 rounded-lg border border-destructive/25 bg-destructive/5 p-3 text-sm text-destructive"><CircleAlertIcon className="mt-0.5 size-4" />{error.message}</div>}{!isLoading && !error && models.length === 0 && <div className="flex h-52 flex-col items-center justify-center gap-2 text-sm text-muted-foreground"><CpuIcon className="size-7" />没有匹配的模型配置</div>}{!isLoading && !error && models.length > 0 && <div className="grid gap-3 p-4 lg:grid-cols-2">{models.map((model) => <article key={model.name} className="rounded-lg border border-border bg-background p-4 transition-colors hover:border-primary/35"><div className="flex items-start justify-between gap-3"><div className="min-w-0"><div className="flex flex-wrap items-center gap-2"><h2 className="truncate font-semibold">{model.display_name || model.name}</h2><Tag theme="success" variant="light" size="small">config.yaml</Tag></div><p className="mt-1 truncate font-mono text-xs text-muted-foreground">{model.name}</p></div><TButton theme="default" variant="outline" size="small" onClick={() => setEditor(model)}>配置</TButton></div>{model.description ? <p className="mt-3 line-clamp-2 min-h-10 text-sm leading-5 text-muted-foreground">{model.description}</p> : <p className="mt-3 min-h-10 text-sm leading-5 text-muted-foreground">未填写描述</p>}<div className="mt-3 grid gap-2 border-t pt-3 text-xs text-muted-foreground sm:grid-cols-2"><span className="truncate" title={model.use}>驱动：{model.use}</span><span className="truncate" title={model.model}>模型：{model.model}</span></div><div className="mt-3 flex flex-wrap gap-1.5">{model.api_key_configured && <Tag size="small" variant="light"><KeyRoundIcon className="mr-1 size-3" />已配置密钥</Tag>}{model.supports_thinking && <Tag theme="primary" size="small" variant="light"><SparklesIcon className="mr-1 size-3" />思考</Tag>}{model.supports_vision && <Tag theme="warning" size="small" variant="light"><EyeIcon className="mr-1 size-3" />视觉</Tag>}{model.supports_reasoning_effort && <Tag size="small" variant="light">推理档位</Tag>}</div></article>)}</div>}</section>{data?.config_path && <p className="text-xs text-muted-foreground">当前配置文件：<code>{data.config_path}</code>{data.updated_at ? ` · 最近更新 ${new Date(data.updated_at).toLocaleString()}` : ""}</p>}</div></main></WorkspaceBody>{editor !== null && <ModelEditor key={editor === "new" ? "new" : editor.name} model={editor === "new" ? null : editor} onClose={() => setEditor(null)} />}</WorkspaceContainer>
|
||||
);
|
||||
}
|
||||
@@ -43,6 +43,7 @@ import MemoryPage from './MemoryPage';
|
||||
import PromptPrefixSettingsPage from './PromptPrefixSettingsPage';
|
||||
import PrivateSquareSettingsPage from './PrivateSquareSettingsPage';
|
||||
import ModelConcurrencySettingsPage from './ModelConcurrencySettingsPage';
|
||||
import ModelManagementPage from './ModelManagementPage';
|
||||
import PublishSquaresSettingsPage from './PublishSquaresSettingsPage';
|
||||
import SkillCompressionSettingsPage from './SkillCompressionSettingsPage';
|
||||
import SkillEvolutionPage from './SkillEvolutionPage';
|
||||
@@ -786,6 +787,18 @@ export default function WorkspaceRoutes({
|
||||
)
|
||||
}
|
||||
/>
|
||||
<Route
|
||||
path="admin/models"
|
||||
element={
|
||||
getStoredAuth()?.system_role === 'admin' ? (
|
||||
<WorkspaceLayout {...layoutProps}>
|
||||
<ModelManagementPage />
|
||||
</WorkspaceLayout>
|
||||
) : (
|
||||
<Navigate to="/page/workspace/chats/new" replace />
|
||||
)
|
||||
}
|
||||
/>
|
||||
<Route
|
||||
path="settings/model-concurrency"
|
||||
element={
|
||||
|
||||
@@ -47,8 +47,9 @@ from app.gateway.routers import (
|
||||
llm_metrics,
|
||||
mcp,
|
||||
memory,
|
||||
menu_overrides,
|
||||
models,
|
||||
menu_overrides,
|
||||
model_management,
|
||||
models,
|
||||
multi_agent,
|
||||
notifications,
|
||||
open_3qfx,
|
||||
@@ -1425,8 +1426,10 @@ This gateway provides custom endpoints for models, MCP configuration, skills, an
|
||||
)
|
||||
|
||||
# Include routers
|
||||
# Models API is mounted at /api/models
|
||||
app.include_router(models.router)
|
||||
# Models API is mounted at /api/models
|
||||
app.include_router(models.router)
|
||||
# Administrator-managed config.yaml model settings and connection tests.
|
||||
app.include_router(model_management.router)
|
||||
|
||||
# MCP API is mounted at /api/mcp
|
||||
app.include_router(mcp.router)
|
||||
|
||||
@@ -0,0 +1,437 @@
|
||||
"""Administrator model configuration backed by DeerFlow's ``config.yaml``.
|
||||
|
||||
The public ``/api/models`` endpoint intentionally stays small and never
|
||||
returns secrets. This router is the complementary administrator surface: it
|
||||
edits only the top-level ``models:`` section in the active DeerFlow config,
|
||||
keeps a disk backup before every write, and swaps the gateway's runtime config
|
||||
after validation so new requests can use the change without a restart.
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import asyncio
|
||||
import copy
|
||||
import logging
|
||||
import os
|
||||
import re
|
||||
import shutil
|
||||
import tempfile
|
||||
from datetime import UTC, datetime
|
||||
from pathlib import Path
|
||||
from time import perf_counter
|
||||
from typing import Any
|
||||
|
||||
import yaml
|
||||
from fastapi import APIRouter, HTTPException, Request
|
||||
from langchain_core.messages import HumanMessage
|
||||
from pydantic import BaseModel, ConfigDict, Field, field_validator
|
||||
|
||||
from app.gateway.deps import get_optional_user_from_request
|
||||
from deerflow.config.app_config import AppConfig, reload_app_config
|
||||
from deerflow.config.model_config import ModelConfig
|
||||
from deerflow.models.factory import create_chat_model
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
router = APIRouter(prefix="/api/admin/models", tags=["admin-models"])
|
||||
|
||||
_MODEL_BLOCK_RE = re.compile(r"(?m)^models:[^\r\n]*(?:\r?\n|$)")
|
||||
_TOP_LEVEL_KEY_RE = re.compile(r"(?m)^[A-Za-z_][A-Za-z0-9_-]*:[^\r\n]*(?:\r?\n|$)")
|
||||
_BACKUP_DIR_NAME = ".model-config-backups"
|
||||
_MUTATION_LOCK = asyncio.Lock()
|
||||
|
||||
_EDITOR_FIELDS = frozenset(
|
||||
{
|
||||
"name",
|
||||
"display_name",
|
||||
"description",
|
||||
"use",
|
||||
"model",
|
||||
"api_key",
|
||||
"base_url",
|
||||
"request_timeout",
|
||||
"max_retries",
|
||||
"max_tokens",
|
||||
"temperature",
|
||||
"stream_usage",
|
||||
"supports_thinking",
|
||||
"supports_reasoning_effort",
|
||||
"supports_vision",
|
||||
"when_thinking_enabled",
|
||||
"when_thinking_disabled",
|
||||
"thinking",
|
||||
"use_responses_api",
|
||||
"output_version",
|
||||
}
|
||||
)
|
||||
_PRESERVED_FIELDS = frozenset(ModelConfig.model_fields) | _EDITOR_FIELDS
|
||||
|
||||
|
||||
class ModelAdminItem(BaseModel):
|
||||
"""A model configuration safe to return to an administrator browser."""
|
||||
|
||||
name: str
|
||||
display_name: str | None = None
|
||||
description: str | None = None
|
||||
use: str
|
||||
model: str
|
||||
base_url: str | None = None
|
||||
request_timeout: float | None = None
|
||||
max_retries: int | None = None
|
||||
max_tokens: int | None = None
|
||||
temperature: float | None = None
|
||||
stream_usage: bool | None = None
|
||||
supports_thinking: bool = False
|
||||
supports_reasoning_effort: bool = False
|
||||
supports_vision: bool = False
|
||||
when_thinking_enabled: dict[str, Any] | None = None
|
||||
when_thinking_disabled: dict[str, Any] | None = None
|
||||
thinking: dict[str, Any] | None = None
|
||||
use_responses_api: bool | None = None
|
||||
output_version: str | None = None
|
||||
extra_settings: dict[str, Any] = Field(default_factory=dict)
|
||||
api_key_configured: bool = False
|
||||
api_key_source: str | None = None
|
||||
source: str = "config.yaml"
|
||||
|
||||
|
||||
class ModelAdminListResponse(BaseModel):
|
||||
models: list[ModelAdminItem]
|
||||
config_path: str
|
||||
updated_at: str | None = None
|
||||
|
||||
|
||||
class ModelUpsertRequest(BaseModel):
|
||||
"""Editable subset of the ``ModelConfig`` schema plus safe extra fields."""
|
||||
|
||||
model_config = ConfigDict(extra="forbid")
|
||||
|
||||
name: str = Field(min_length=1, max_length=160)
|
||||
display_name: str | None = Field(default=None, max_length=160)
|
||||
description: str | None = Field(default=None, max_length=1000)
|
||||
use: str = Field(min_length=1, max_length=255)
|
||||
model: str = Field(min_length=1, max_length=255)
|
||||
# ``None`` means "keep the currently configured secret" on update.
|
||||
# The value is write-only in practice: it is never part of a response.
|
||||
api_key: str | None = Field(default=None, max_length=4096)
|
||||
clear_api_key: bool = False
|
||||
base_url: str | None = Field(default=None, max_length=2048)
|
||||
request_timeout: float | None = Field(default=None, ge=1, le=600)
|
||||
max_retries: int | None = Field(default=None, ge=0, le=20)
|
||||
max_tokens: int | None = Field(default=None, ge=1, le=1_000_000)
|
||||
temperature: float | None = Field(default=None, ge=0, le=2)
|
||||
stream_usage: bool | None = None
|
||||
supports_thinking: bool = False
|
||||
supports_reasoning_effort: bool = False
|
||||
supports_vision: bool = False
|
||||
when_thinking_enabled: dict[str, Any] | None = None
|
||||
when_thinking_disabled: dict[str, Any] | None = None
|
||||
thinking: dict[str, Any] | None = None
|
||||
use_responses_api: bool | None = None
|
||||
output_version: str | None = Field(default=None, max_length=128)
|
||||
extra_settings: dict[str, Any] = Field(default_factory=dict)
|
||||
|
||||
@field_validator("name", "use", "model")
|
||||
@classmethod
|
||||
def _trim_required(cls, value: str) -> str:
|
||||
normalized = value.strip()
|
||||
if not normalized:
|
||||
raise ValueError("value cannot be empty")
|
||||
return normalized
|
||||
|
||||
@field_validator("display_name", "description", "base_url", "api_key", "output_version")
|
||||
@classmethod
|
||||
def _trim_optional(cls, value: str | None) -> str | None:
|
||||
return value.strip() if isinstance(value, str) else value
|
||||
|
||||
|
||||
class ModelTestRequest(BaseModel):
|
||||
model: ModelUpsertRequest
|
||||
|
||||
|
||||
class ModelTestResponse(BaseModel):
|
||||
success: bool
|
||||
latency_ms: int
|
||||
message: str
|
||||
preview: str | None = None
|
||||
|
||||
|
||||
def _mask_env_source(value: Any) -> str | None:
|
||||
if isinstance(value, str) and value.startswith("$") and len(value) > 1:
|
||||
return value[1:]
|
||||
return None
|
||||
|
||||
|
||||
def _to_admin_item(raw: dict[str, Any]) -> ModelAdminItem:
|
||||
api_key = raw.get("api_key")
|
||||
extra = {
|
||||
key: copy.deepcopy(value)
|
||||
for key, value in raw.items()
|
||||
if key not in _PRESERVED_FIELDS
|
||||
}
|
||||
return ModelAdminItem(
|
||||
name=str(raw.get("name") or ""),
|
||||
display_name=raw.get("display_name"),
|
||||
description=raw.get("description"),
|
||||
use=str(raw.get("use") or ""),
|
||||
model=str(raw.get("model") or ""),
|
||||
base_url=raw.get("base_url"),
|
||||
request_timeout=raw.get("request_timeout"),
|
||||
max_retries=raw.get("max_retries"),
|
||||
max_tokens=raw.get("max_tokens"),
|
||||
temperature=raw.get("temperature"),
|
||||
stream_usage=raw.get("stream_usage"),
|
||||
supports_thinking=bool(raw.get("supports_thinking", False)),
|
||||
supports_reasoning_effort=bool(raw.get("supports_reasoning_effort", False)),
|
||||
supports_vision=bool(raw.get("supports_vision", False)),
|
||||
when_thinking_enabled=copy.deepcopy(raw.get("when_thinking_enabled")),
|
||||
when_thinking_disabled=copy.deepcopy(raw.get("when_thinking_disabled")),
|
||||
thinking=copy.deepcopy(raw.get("thinking")),
|
||||
use_responses_api=raw.get("use_responses_api"),
|
||||
output_version=raw.get("output_version"),
|
||||
extra_settings=extra,
|
||||
api_key_configured=bool(api_key),
|
||||
api_key_source=_mask_env_source(api_key),
|
||||
)
|
||||
|
||||
|
||||
def _resolve_config_path() -> Path:
|
||||
return AppConfig.resolve_config_path()
|
||||
|
||||
|
||||
def _read_config_data(path: Path) -> tuple[str, dict[str, Any], list[dict[str, Any]]]:
|
||||
raw_text = path.read_text(encoding="utf-8")
|
||||
try:
|
||||
data = yaml.safe_load(raw_text) or {}
|
||||
except yaml.YAMLError as exc:
|
||||
raise HTTPException(status_code=500, detail=f"Unable to parse active config.yaml: {exc}") from exc
|
||||
if not isinstance(data, dict):
|
||||
raise HTTPException(status_code=500, detail="The active config.yaml root must be a mapping")
|
||||
raw_models = data.get("models") or []
|
||||
if not isinstance(raw_models, list) or not all(isinstance(item, dict) for item in raw_models):
|
||||
raise HTTPException(status_code=500, detail="config.yaml models must be a list of mappings")
|
||||
return raw_text, data, [copy.deepcopy(item) for item in raw_models]
|
||||
|
||||
|
||||
def _render_models_block(models: list[dict[str, Any]], newline: str) -> str:
|
||||
rendered = yaml.safe_dump(
|
||||
{"models": models},
|
||||
allow_unicode=True,
|
||||
sort_keys=False,
|
||||
default_flow_style=False,
|
||||
width=120,
|
||||
)
|
||||
if newline != "\n":
|
||||
rendered = rendered.replace("\n", newline)
|
||||
return rendered if rendered.endswith(newline) else f"{rendered}{newline}"
|
||||
|
||||
|
||||
def _replace_models_block(raw_text: str, models: list[dict[str, Any]]) -> str:
|
||||
"""Replace the active model entries without touching other config sections.
|
||||
|
||||
PyYAML does not retain comments. Keep every comment found in the old
|
||||
model section as documentation after the newly rendered model list rather
|
||||
than silently deleting an operator's notes or disabled examples.
|
||||
"""
|
||||
|
||||
block_match = _MODEL_BLOCK_RE.search(raw_text)
|
||||
if block_match is None:
|
||||
raise HTTPException(status_code=500, detail="The active config.yaml has no top-level models section")
|
||||
newline = "\r\n" if "\r\n" in raw_text else "\n"
|
||||
next_key = _TOP_LEVEL_KEY_RE.search(raw_text, block_match.end())
|
||||
end = next_key.start() if next_key else len(raw_text)
|
||||
previous_section = raw_text[block_match.end() : end]
|
||||
preserved_comments = "".join(
|
||||
line
|
||||
for line in previous_section.splitlines(keepends=True)
|
||||
if line.lstrip().startswith("#")
|
||||
)
|
||||
rendered = _render_models_block(models, newline)
|
||||
if preserved_comments:
|
||||
if not preserved_comments.endswith(("\n", "\r")):
|
||||
preserved_comments = f"{preserved_comments}{newline}"
|
||||
rendered = f"{rendered}{preserved_comments}"
|
||||
return f"{raw_text[:block_match.start()]}{rendered}{raw_text[end:]}"
|
||||
|
||||
|
||||
def _backup_and_write_config(path: Path, raw_before: str, raw_after: str) -> Path:
|
||||
backup_dir = path.parent / _BACKUP_DIR_NAME
|
||||
backup_dir.mkdir(parents=True, exist_ok=True)
|
||||
stamp = datetime.now(UTC).strftime("%Y%m%dT%H%M%S%fZ")
|
||||
backup_path = backup_dir / f"{path.stem}.{stamp}.yaml"
|
||||
# Persist the exact text that was read for this mutation. This makes the
|
||||
# rollback deterministic even if another operator changes the file while
|
||||
# the gateway is processing a request.
|
||||
with backup_path.open("w", encoding="utf-8", newline="") as handle:
|
||||
handle.write(raw_before)
|
||||
|
||||
fd, temp_name = tempfile.mkstemp(prefix=f".{path.name}.", suffix=".tmp", dir=path.parent)
|
||||
try:
|
||||
with os.fdopen(fd, "w", encoding="utf-8", newline="") as handle:
|
||||
handle.write(raw_after)
|
||||
handle.flush()
|
||||
os.fsync(handle.fileno())
|
||||
os.replace(temp_name, path)
|
||||
except Exception:
|
||||
try:
|
||||
os.unlink(temp_name)
|
||||
except OSError:
|
||||
pass
|
||||
raise
|
||||
return backup_path
|
||||
|
||||
|
||||
def _restore_backup(path: Path, backup_path: Path) -> None:
|
||||
try:
|
||||
shutil.copy2(backup_path, path)
|
||||
except Exception:
|
||||
logger.exception("Failed to restore model config backup at %s", backup_path)
|
||||
|
||||
|
||||
def _request_to_raw_model(payload: ModelUpsertRequest, existing: dict[str, Any] | None = None) -> dict[str, Any]:
|
||||
merged = copy.deepcopy(existing or {})
|
||||
supplied = payload.model_dump(exclude={"extra_settings", "clear_api_key"}, exclude_none=True)
|
||||
merged.update(supplied)
|
||||
for key, value in payload.extra_settings.items():
|
||||
normalized = str(key).strip()
|
||||
if not normalized:
|
||||
raise HTTPException(status_code=422, detail="Extra setting names cannot be blank")
|
||||
if normalized in _EDITOR_FIELDS or normalized in {"api_key", "clear_api_key", "extra_settings"}:
|
||||
raise HTTPException(status_code=422, detail=f"Extra setting '{normalized}' duplicates a managed field")
|
||||
merged[normalized] = copy.deepcopy(value)
|
||||
if payload.clear_api_key:
|
||||
merged.pop("api_key", None)
|
||||
elif payload.api_key is None and existing is not None and "api_key" in existing:
|
||||
merged["api_key"] = existing["api_key"]
|
||||
|
||||
# Validate the raw schema without resolving `$ENV_VAR` values. Environment
|
||||
# resolution happens during the post-write runtime reload and is rolled back
|
||||
# if unavailable.
|
||||
try:
|
||||
ModelConfig.model_validate(merged)
|
||||
except Exception as exc:
|
||||
raise HTTPException(status_code=422, detail=f"Invalid model configuration: {exc}") from exc
|
||||
return merged
|
||||
|
||||
|
||||
def _redact_text(value: str, secrets: list[str]) -> str:
|
||||
result = value
|
||||
for secret in secrets:
|
||||
if secret:
|
||||
result = result.replace(secret, "***")
|
||||
return result[:800]
|
||||
|
||||
|
||||
async def _require_admin(request: Request) -> None:
|
||||
user = getattr(request.state, "user", None)
|
||||
if user is None:
|
||||
user = await get_optional_user_from_request(request)
|
||||
if user is None:
|
||||
raise HTTPException(status_code=401, detail="Authentication is required")
|
||||
if getattr(user, "system_role", None) != "admin":
|
||||
raise HTTPException(status_code=403, detail="Model configuration requires an administrator")
|
||||
|
||||
|
||||
async def _save_models(request: Request, raw_before: str, models: list[dict[str, Any]]) -> list[dict[str, Any]]:
|
||||
path = _resolve_config_path()
|
||||
raw_after = _replace_models_block(raw_before, models)
|
||||
backup_path = await asyncio.to_thread(_backup_and_write_config, path, raw_before, raw_after)
|
||||
try:
|
||||
config = reload_app_config(str(path))
|
||||
request.app.state.config = config
|
||||
except Exception as exc:
|
||||
logger.warning("Updated model configuration could not be activated: %s", type(exc).__name__)
|
||||
await asyncio.to_thread(_restore_backup, path, backup_path)
|
||||
try:
|
||||
request.app.state.config = reload_app_config(str(path))
|
||||
except Exception:
|
||||
logger.exception("Failed to reload restored model config")
|
||||
raise HTTPException(
|
||||
status_code=422,
|
||||
detail="The model configuration could not be activated; the previous config was restored.",
|
||||
) from exc
|
||||
return models
|
||||
|
||||
|
||||
@router.get("", response_model=ModelAdminListResponse)
|
||||
async def list_model_configurations(request: Request) -> ModelAdminListResponse:
|
||||
await _require_admin(request)
|
||||
path = _resolve_config_path()
|
||||
_, _, models = await asyncio.to_thread(_read_config_data, path)
|
||||
updated_at = datetime.fromtimestamp(path.stat().st_mtime, tz=UTC).isoformat()
|
||||
return ModelAdminListResponse(
|
||||
models=[_to_admin_item(item) for item in models],
|
||||
config_path=str(path),
|
||||
updated_at=updated_at,
|
||||
)
|
||||
|
||||
|
||||
@router.post("", response_model=ModelAdminItem, status_code=201)
|
||||
async def create_model_configuration(body: ModelUpsertRequest, request: Request) -> ModelAdminItem:
|
||||
await _require_admin(request)
|
||||
async with _MUTATION_LOCK:
|
||||
path = _resolve_config_path()
|
||||
raw_before, _, models = await asyncio.to_thread(_read_config_data, path)
|
||||
if any(str(item.get("name")) == body.name for item in models):
|
||||
raise HTTPException(status_code=409, detail=f"A model named '{body.name}' already exists")
|
||||
created = _request_to_raw_model(body)
|
||||
await _save_models(request, raw_before, [*models, created])
|
||||
return _to_admin_item(created)
|
||||
|
||||
|
||||
@router.put("/{model_name}", response_model=ModelAdminItem)
|
||||
async def update_model_configuration(model_name: str, body: ModelUpsertRequest, request: Request) -> ModelAdminItem:
|
||||
await _require_admin(request)
|
||||
if body.name != model_name:
|
||||
raise HTTPException(status_code=422, detail="A model's stable name cannot be changed; edit its display name instead")
|
||||
async with _MUTATION_LOCK:
|
||||
path = _resolve_config_path()
|
||||
raw_before, _, models = await asyncio.to_thread(_read_config_data, path)
|
||||
index = next((i for i, item in enumerate(models) if str(item.get("name")) == model_name), None)
|
||||
if index is None:
|
||||
raise HTTPException(status_code=404, detail=f"Model '{model_name}' was not found")
|
||||
updated = _request_to_raw_model(body, models[index])
|
||||
next_models = [*models]
|
||||
next_models[index] = updated
|
||||
await _save_models(request, raw_before, next_models)
|
||||
return _to_admin_item(updated)
|
||||
|
||||
|
||||
@router.post("/test", response_model=ModelTestResponse)
|
||||
async def test_model_connection(body: ModelTestRequest, request: Request) -> ModelTestResponse:
|
||||
"""Perform one bounded minimal completion without persisting the draft."""
|
||||
|
||||
await _require_admin(request)
|
||||
path = _resolve_config_path()
|
||||
_, _, models = await asyncio.to_thread(_read_config_data, path)
|
||||
existing = next((item for item in models if str(item.get("name")) == body.model.name), None)
|
||||
candidate = _request_to_raw_model(body.model, existing)
|
||||
candidate_secret = str(candidate.get("api_key") or "")
|
||||
secrets = [candidate_secret]
|
||||
try:
|
||||
runtime_candidate = AppConfig.resolve_env_variables(copy.deepcopy(candidate))
|
||||
# An environment-variable API key is resolved only in memory for this
|
||||
# request. Redact both forms because provider exceptions can include
|
||||
# either the original `$NAME` reference or its resolved value.
|
||||
resolved_secret = str(runtime_candidate.get("api_key") or "")
|
||||
secrets = [candidate_secret, resolved_secret]
|
||||
model_config = ModelConfig.model_validate(runtime_candidate)
|
||||
base_config = request.app.state.config
|
||||
test_config = base_config.model_copy(update={"models": [model_config]})
|
||||
chat_model = create_chat_model(model_config.name, app_config=test_config, force_disable_thinking=True)
|
||||
started = perf_counter()
|
||||
response = await asyncio.wait_for(
|
||||
chat_model.ainvoke([HumanMessage(content="Reply with exactly: OK")]),
|
||||
timeout=min(float(candidate.get("request_timeout") or 30), 45),
|
||||
)
|
||||
latency_ms = round((perf_counter() - started) * 1000)
|
||||
preview = _redact_text(str(getattr(response, "content", "OK")), secrets)
|
||||
return ModelTestResponse(success=True, latency_ms=latency_ms, message="Connection successful", preview=preview)
|
||||
except Exception as exc:
|
||||
# Provider errors sometimes echo a credential. Never log the full
|
||||
# exception or return it without redacting the submitted key.
|
||||
logger.warning("Model connection test failed for '%s': %s", candidate.get("name"), type(exc).__name__)
|
||||
return ModelTestResponse(
|
||||
success=False,
|
||||
latency_ms=0,
|
||||
message=_redact_text(str(exc) or type(exc).__name__, secrets),
|
||||
)
|
||||
@@ -0,0 +1,92 @@
|
||||
"""Focused regression tests for administrator model configuration helpers."""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
from types import SimpleNamespace
|
||||
|
||||
import pytest
|
||||
import yaml
|
||||
|
||||
from app.gateway.routers import model_management as router
|
||||
|
||||
|
||||
def test_replace_models_block_keeps_other_sections_and_comments() -> None:
|
||||
raw = """config_version: 1
|
||||
models:
|
||||
# Existing provider note
|
||||
- name: old-model
|
||||
use: langchain_openai:ChatOpenAI
|
||||
model: old-model
|
||||
# Disabled provider example remains useful to operators.
|
||||
# - name: disabled-model
|
||||
# model: disabled-model
|
||||
agents:
|
||||
default: lead_agent
|
||||
"""
|
||||
|
||||
replaced = router._replace_models_block(
|
||||
raw,
|
||||
[
|
||||
{
|
||||
"name": "new-model",
|
||||
"use": "langchain_openai:ChatOpenAI",
|
||||
"model": "new-model",
|
||||
}
|
||||
],
|
||||
)
|
||||
|
||||
parsed = yaml.safe_load(replaced)
|
||||
assert parsed["models"] == [
|
||||
{
|
||||
"name": "new-model",
|
||||
"use": "langchain_openai:ChatOpenAI",
|
||||
"model": "new-model",
|
||||
}
|
||||
]
|
||||
assert parsed["agents"] == {"default": "lead_agent"}
|
||||
assert "# Existing provider note" in replaced
|
||||
assert "# Disabled provider example remains useful to operators." in replaced
|
||||
|
||||
|
||||
def test_connection_error_redaction_hides_literal_and_resolved_keys() -> None:
|
||||
message = "provider rejected raw-secret and resolved-secret"
|
||||
assert router._redact_text(message, ["raw-secret", "resolved-secret"]) == "provider rejected *** and ***"
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_connection_test_uses_one_minimal_prompt_without_writing_config(
|
||||
tmp_path,
|
||||
monkeypatch: pytest.MonkeyPatch,
|
||||
) -> None:
|
||||
config_path = tmp_path / "config.yaml"
|
||||
original = "models: []\nagent: lead_agent\n"
|
||||
config_path.write_text(original, encoding="utf-8")
|
||||
messages_seen = []
|
||||
|
||||
class FakeChatModel:
|
||||
async def ainvoke(self, messages):
|
||||
messages_seen.extend(messages)
|
||||
return SimpleNamespace(content="OK")
|
||||
|
||||
monkeypatch.setattr(router, "_resolve_config_path", lambda: config_path)
|
||||
monkeypatch.setattr(router, "create_chat_model", lambda *_args, **_kwargs: FakeChatModel())
|
||||
request = SimpleNamespace(
|
||||
state=SimpleNamespace(user=SimpleNamespace(system_role="admin")),
|
||||
app=SimpleNamespace(state=SimpleNamespace(config=SimpleNamespace(model_copy=lambda **_kwargs: SimpleNamespace()))),
|
||||
)
|
||||
body = router.ModelTestRequest(
|
||||
model=router.ModelUpsertRequest(
|
||||
name="test-model",
|
||||
use="langchain_openai:ChatOpenAI",
|
||||
model="test-model",
|
||||
api_key="raw-secret",
|
||||
)
|
||||
)
|
||||
|
||||
result = await router.test_model_connection(body, request)
|
||||
|
||||
assert result.success is True
|
||||
assert result.preview == "OK"
|
||||
assert len(messages_seen) == 1
|
||||
assert messages_seen[0].content == "Reply with exactly: OK"
|
||||
assert config_path.read_text(encoding="utf-8") == original
|
||||
Reference in new issue
Block a user