From af398f22e12fdc3ad87ff92adb63035e4ea98e25 Mon Sep 17 00:00:00 2001
From: Liu Xiang Qian <smartlitchi@gmail.com>
Date: Wed, 5 Nov 2025 09:31:58 +0800
Subject: [PATCH] fix: add AI_MAX_TOKENS environment variable to prevent
 response truncation
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

## Problem
AI responses were being truncated due to a hardcoded max_tokens limit of 2000,
causing JSON parsing failures. The error occurred when:
1. AI's thought process analysis was cut off mid-response
2. extractDecisions() incorrectly extracted MACD data arrays from the input prompt
3. Go failed to unmarshal numbers into Decision struct

Error message:
```
json: cannot unmarshal number into Go value of type decision.Decision
JSON内容: [-867.759, -937.406, -1020.435, ...]
```

## Solution
- Add MaxTokens field to mcp.Client struct
- Read AI_MAX_TOKENS from environment variable (default: 2000)
- Set AI_MAX_TOKENS=4000 in docker-compose.yml for production use
- This provides enough tokens for complete analysis with the 800-line trading strategy prompt

## Testing
- Verify environment variable is read correctly
- Confirm AI responses are no longer truncated
- Check decision logs for complete JSON output
---
 docker-compose.yml |  1 +
 mcp/client.go      | 25 ++++++++++++++++++++-----
 2 files changed, 21 insertions(+), 5 deletions(-)

diff --git a/docker-compose.yml b/docker-compose.yml
index a9d35026..dc25bb44 100644
--- a/docker-compose.yml
+++ b/docker-compose.yml
@@ -17,6 +17,7 @@ services:
       - /etc/localtime:/etc/localtime:ro  # Sync host time
     environment:
       - TZ=${NOFX_TIMEZONE:-Asia/Shanghai}  # Set timezone
+      - AI_MAX_TOKENS=4000  # AI响应的最大token数（默认2000，建议4000-8000）
     networks:
       - nofx-network
     healthcheck:
diff --git a/mcp/client.go b/mcp/client.go
index 9191dfaf..dd0f873a 100644
--- a/mcp/client.go
+++ b/mcp/client.go
@@ -7,6 +7,8 @@ import (
 	"io"
 	"log"
 	"net/http"
+	"os"
+	"strconv"
 	"strings"
 	"time"
 )
@@ -28,15 +30,28 @@ type Client struct {
 	Model      string
 	Timeout    time.Duration
 	UseFullURL bool // 是否使用完整URL（不添加/chat/completions）
+	MaxTokens  int  // AI响应的最大token数
 }
 
 func New() *Client {
+	// 从环境变量读取 MaxTokens，默认 2000
+	maxTokens := 2000
+	if envMaxTokens := os.Getenv("AI_MAX_TOKENS"); envMaxTokens != "" {
+		if parsed, err := strconv.Atoi(envMaxTokens); err == nil && parsed > 0 {
+			maxTokens = parsed
+			log.Printf("🔧 [MCP] 使用环境变量 AI_MAX_TOKENS: %d", maxTokens)
+		} else {
+			log.Printf("⚠️  [MCP] 环境变量 AI_MAX_TOKENS 无效 (%s)，使用默认值: %d", envMaxTokens, maxTokens)
+		}
+	}
+
 	// 默认配置
 	return &Client{
-		Provider: ProviderDeepSeek,
-		BaseURL:  "https://api.deepseek.com/v1",
-		Model:    "deepseek-chat",
-		Timeout:  120 * time.Second, // 增加到120秒，因为AI需要分析大量数据
+		Provider:  ProviderDeepSeek,
+		BaseURL:   "https://api.deepseek.com/v1",
+		Model:     "deepseek-chat",
+		Timeout:   120 * time.Second, // 增加到120秒，因为AI需要分析大量数据
+		MaxTokens: maxTokens,
 	}
 }
 
@@ -190,7 +205,7 @@ func (client *Client) callOnce(systemPrompt, userPrompt string) (string, error)
 		"model":       client.Model,
 		"messages":    messages,
 		"temperature": 0.5, // 降低temperature以提高JSON格式稳定性
-		"max_tokens":  2000,
+		"max_tokens":  client.MaxTokens,
 	}
 
 	// 注意：response_format 参数仅 OpenAI 支持，DeepSeek/Qwen 不支持