Research Agent MCP
by myeong-ga
use-anthropic-chat.ts•11 kB
"use client"
import { useState, useEffect, useCallback } from "react"
import type { CreateMessage, Message } from "@ai-sdk/react"
import type { Source, ChatStatus, TokenUsage, ModelConfig } from "@/lib/types"
import { nanoid } from "@/lib/nanoid"
import { useLlmProvider } from "@/contexts/llm-provider-context"
import type { SearchSuggestion } from "@/lib/types"
import { getDefaultModelConfig } from "@/lib/models"
import { toast } from "sonner"
export function useAnthropicChat() {
const { getSelectedModel ,getReasoningType} = useLlmProvider()
const [messages, setMessages] = useState<Message[]>([])
const [sources, setSources] = useState<Source[]>([])
const [searchSuggestions, setSearchSuggestions] = useState<SearchSuggestion[]>([])
const [searchSuggestionsReasoning, setSearchSuggestionsReasoning] = useState<string>("")
const [searchSuggestionsConfidence, setSearchSuggestionsConfidence] = useState<number | null>(null)
const [status, setStatus] = useState<ChatStatus>("ready")
const [error, setError] = useState<string | null>(null)
const [controller, setController] = useState<AbortController | null>(null)
const [chatId, setChatId] = useState<string>(() => nanoid())
const [tokenUsage, setTokenUsage] = useState<TokenUsage | null>(null)
const [modelConfig, setModelConfig] = useState<ModelConfig>(() => {
const selectedModelId = getSelectedModel("anthropic")
return getDefaultModelConfig(selectedModelId).config
})
const [responseModelConfig, setResponseModelConfig] = useState<ModelConfig | null>(null) // 마지막 응답에서 확인할 수 있는 model config , chatting 창 message display 와 함께 사용된다
const [responseSelectModel, setResponseSelectModel] = useState<string | null>(null) // 마지막 응답에서 확인할 수 있는 model config , chatting 창 message display 와 함께 사용된다
const [responseReasoningType, setResponseReasoningType] = useState<string | null>(null) // 마지막 응답에서 확인할 수 있는 model config , chatting 창 message display 와 함께 사용된다
const selectedModelId = getSelectedModel("anthropic")
const reasoningType = getReasoningType("anthropic", selectedModelId)
// Update model config when selected model changes
useEffect(() => {
const selectedModelId = getSelectedModel("anthropic")
const defaultConfig = getDefaultModelConfig(selectedModelId).config
setModelConfig(defaultConfig)
}, [getSelectedModel])
useEffect(() => {
if (error && messages.length > 0) {
setError(null)
}
}, [messages, error])
useEffect(() => {
return () => {
if (controller) {
controller.abort()
}
}
}, [controller])
const updateModelConfig = useCallback((config: Partial<ModelConfig>, showToast = false) => {
setModelConfig((prevConfig) => {
const newConfig = {
...prevConfig,
...config,
}
if (showToast) {
toast.success("Model configuration updated", {
description: `Temperature: ${newConfig.temperature.toFixed(2)}, Max Tokens: ${newConfig.maxTokens}`,
duration: 3000,
})
}
return newConfig
})
}, [])
const stop = useCallback(() => {
if (controller) {
controller.abort()
setController(null)
setStatus("ready")
}
}, [controller])
const resetChat = useCallback(() => {
stop()
setMessages([])
setSources([])
setSearchSuggestions([])
setSearchSuggestionsReasoning("")
setSearchSuggestionsConfidence(null)
setTokenUsage(null)
setResponseModelConfig(null)
setError(null)
setChatId(nanoid())
// Reset model config to default
const selectedModelId = getSelectedModel("anthropic")
setModelConfig(getDefaultModelConfig(selectedModelId).config)
}, [stop, getSelectedModel])
const sendMessage = async (message: string | CreateMessage) => {
try {
if (controller) {
controller.abort()
}
setSources([])
setSearchSuggestions([])
setSearchSuggestionsReasoning("")
setSearchSuggestionsConfidence(null)
setTokenUsage(null)
setResponseModelConfig(null)
setError(null)
const newController = new AbortController()
setController(newController)
const userMessage: Message =
typeof message === "string" ? { id: nanoid(), role: "user", content: message } : { id: nanoid(), ...message }
setStatus("submitted")
setMessages((prevMessages) => [...prevMessages, userMessage])
const response = await fetch("/api/chat/anthropic", {
method: "POST",
headers: {
"Content-Type": "application/json",
},
body: JSON.stringify({
messages: [...messages, userMessage],
model: selectedModelId,
chatId,
modelConfig,
reasoningType
}),
signal: newController.signal,
})
if (!response.ok) {
throw new Error(`HTTP error! status: ${response.status}`)
}
if (!response.body) {
throw new Error("Response body is null")
}
setStatus("streaming")
const assistantMessageId = nanoid()
setMessages((prevMessages) => [...prevMessages, { id: assistantMessageId, role: "assistant", content: "" }])
const reader = response.body.getReader()
const decoder = new TextDecoder()
let accumulatedText = ""
try {
while (true) {
const { done, value } = await reader.read()
if (done) break
const chunk = decoder.decode(value)
const lines = chunk.split("\n\n")
for (const line of lines) {
if (line.startsWith("data: ")) {
try {
const data = JSON.parse(line.substring(6))
if (data.type === "text-delta") {
accumulatedText += data.text
setMessages((prevMessages) =>
prevMessages.map((msg) =>
msg.id === assistantMessageId ? { ...msg, content: accumulatedText } : msg,
),
)
} else if (data.type === "sources" && Array.isArray(data.sources)) {
setSources(data.sources)
} else if (data.type === "searchSuggestions") {
if (data.searchSuggestions && Array.isArray(data.searchSuggestions)) {
const suggestions = data.searchSuggestions.map((term: string) => ({ term }))
setSearchSuggestions(suggestions)
}
if (data.confidence && typeof data.confidence === "number") {
setSearchSuggestionsConfidence(data.confidence)
}
if (data.reasoning && typeof data.reasoning === "string") {
setSearchSuggestionsReasoning(data.reasoning)
}
} else if (data.type === "model-config" && typeof data.config === "object") {
// Handle model config response from server
const config = data.config as ModelConfig
setResponseModelConfig(config)
} else if (data.type === "selected-model" && typeof data.model === "string") {
const responseModel = data.model as unknown as string
setResponseSelectModel(responseModel)
} else if (data.type === "reasoning-type" && typeof data.reasoning === "string") {
const responseReasoning = data.reasoning as string
setResponseReasoningType(responseReasoning)
} else if (data.type === "cleaned-text" && "text" in data && typeof data.text === "string") {
setMessages((prevMessages) => {
const newMessages = [...prevMessages]
for (let i = newMessages.length - 1; i >= 0; i--) {
if (newMessages[i].role === "assistant") {
newMessages[i] = {
...newMessages[i],
content: data.text as string,
}
break
}
}
return newMessages
})
} else if (data.type === "usage" && typeof data.usage === "object") {
const usage = data.usage
setTokenUsage((prev) =>
prev
? {
...prev,
promptTokens: usage.input_tokens || 0,
completionTokens: usage.output_tokens || 0,
totalTokens: (usage.input_tokens || 0) + (usage.output_tokens || 0),
}
: {
promptTokens: usage.input_tokens || 0,
completionTokens: usage.output_tokens || 0,
totalTokens: (usage.input_tokens || 0) + (usage.output_tokens || 0),
finishReason: "unknown",
},
)
} else if (data.type === "stop_reason") {
setTokenUsage((prev) =>
prev
? {
...prev,
finishReason: data.stop_reason,
}
: {
promptTokens: 0,
completionTokens: 0,
totalTokens: 0,
finishReason: data.stop_reason,
},
)
} else if (data.type === "error") {
setError(data.error)
setStatus("error")
}
} catch (e) {
console.error("Error parsing SSE data:", e)
}
}
}
}
setStatus("ready")
} catch (err) {
console.error("Error reading stream:", err)
setError("Error reading response stream")
setStatus("error")
} finally {
setController(null)
if (status !== "error") {
setStatus("ready")
}
}
} catch (err) {
console.error("Error sending message:", err)
setController(null)
setError(err instanceof Error ? err.message : "Unknown error occurred")
setStatus("error")
}
}
return {
messages,
status,
error,
stop,
sources,
searchSuggestions,
searchSuggestionsReasoning,
searchSuggestionsConfidence,
sendMessage,
chatId,
resetChat,
tokenUsage,
modelConfig,
responseModelConfig,
responseSelectModel,
responseReasoningType,
updateModelConfig,
}
}
Latest Blog Posts
- Federated Learning with MCP: Building Privacy-Preserving Agents Across Distributed EdgesBy Om-Shree-0709 on .SecuremcpLearning
- MCP Moves to the Linux Foundation: Neutral Stewardship for Agentic InfrastructureBy Om-Shree-0709 on .mcpanthropicLinux Foundation
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/myeong-ga/research-agent-mcp-0.36-pro-preview-06-01'
If you have feedback or need assistance with the MCP directory API, please join our Discord server