diff --git a/pom.xml b/pom.xml
index 61a4ec4..7eed89e 100644
--- a/pom.xml
+++ b/pom.xml
@@ -81,11 +81,50 @@
             <version>8.5.7</version>
         </dependency>
 
-        <!-- aws-java 用于存储桶 -->
+        <!-- 文档解析依赖 -->
+        <!-- PDF解析 -->
         <dependency>
-            <groupId>com.amazonaws</groupId>
-            <artifactId>aws-java-sdk-s3</artifactId>
-            <version>1.12.700</version>
+            <groupId>org.apache.pdfbox</groupId>
+            <artifactId>pdfbox</artifactId>
+            <version>2.0.30</version>
+        </dependency>
+
+        <!-- Word文档解析 -->
+        <dependency>
+            <groupId>org.apache.poi</groupId>
+            <artifactId>poi-ooxml</artifactId>
+            <version>5.2.5</version>
+        </dependency>
+        <dependency>
+            <groupId>org.apache.poi</groupId>
+            <artifactId>poi-scratchpad</artifactId>
+            <version>5.2.5</version>
+        </dependency>
+
+        <!-- Excel解析 -->
+        <dependency>
+            <groupId>org.apache.poi</groupId>
+            <artifactId>poi-ooxml-schemas</artifactId>
+            <version>4.1.2</version>
+        </dependency>
+
+        <!-- Markdown解析 -->
+        <dependency>
+            <groupId>com.vladsch.flexmark</groupId>
+            <artifactId>flexmark-all</artifactId>
+            <version>0.64.8</version>
+        </dependency>
+
+        <!-- Tika 用于通用文档类型检测和内容提取 -->
+        <dependency>
+            <groupId>org.apache.tika</groupId>
+            <artifactId>tika-core</artifactId>
+            <version>2.9.1</version>
+        </dependency>
+        <dependency>
+            <groupId>org.apache.tika</groupId>
+            <artifactId>tika-parsers-standard-package</artifactId>
+            <version>2.9.1</version>
         </dependency>
 
         <dependency>
diff --git a/src/main/java/cn/yinlihupo/common/util/DocumentParserUtil.java b/src/main/java/cn/yinlihupo/common/util/DocumentParserUtil.java
new file mode 100644
index 0000000..84360f3
--- /dev/null
+++ b/src/main/java/cn/yinlihupo/common/util/DocumentParserUtil.java
@@ -0,0 +1,261 @@
+package cn.yinlihupo.common.util;
+
+import com.vladsch.flexmark.html.HtmlRenderer;
+import com.vladsch.flexmark.parser.Parser;
+import com.vladsch.flexmark.util.data.MutableDataSet;
+import lombok.extern.slf4j.Slf4j;
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.text.PDFTextStripper;
+import org.apache.poi.hwpf.HWPFDocument;
+import org.apache.poi.hwpf.extractor.WordExtractor;
+import org.apache.poi.xwpf.usermodel.XWPFDocument;
+import org.apache.poi.xwpf.usermodel.XWPFParagraph;
+import org.apache.poi.ss.usermodel.*;
+import org.apache.tika.Tika;
+import org.apache.tika.metadata.Metadata;
+
+import java.io.BufferedInputStream;
+import java.io.ByteArrayInputStream;
+import java.io.IOException;
+import java.io.InputStream;
+import java.util.List;
+
+/**
+ * 文档解析工具类
+ * 支持 PDF、Word、Excel、Markdown 等格式的文档解析
+ */
+@Slf4j
+public class DocumentParserUtil {
+
+    private static final Tika tika = new Tika();
+
+    /**
+     * 自动检测文件类型并解析内容
+     *
+     * @param inputStream 文件输入流
+     * @param fileName    文件名
+     * @return 文档文本内容
+     */
+    public static String parse(InputStream inputStream, String fileName) {
+        try {
+            // 将输入流转换为支持 mark/reset 的 BufferedInputStream
+            BufferedInputStream bufferedStream = new BufferedInputStream(inputStream);
+            bufferedStream.mark(Integer.MAX_VALUE);
+
+            // 检测文件类型
+            String mimeType = tika.detect(bufferedStream, fileName);
+            log.info("检测到文件类型: {}, 文件名: {}", mimeType, fileName);
+
+            // 重置输入流到起始位置
+            bufferedStream.reset();
+
+            // 根据文件扩展名或MIME类型选择解析器
+            String lowerFileName = fileName.toLowerCase();
+
+            if (lowerFileName.endsWith(".pdf") || mimeType.equals("application/pdf")) {
+                return parsePdf(bufferedStream);
+            } else if (lowerFileName.endsWith(".docx") || mimeType.equals("application/vnd.openxmlformats-officedocument.wordprocessingml.document")) {
+                return parseDocx(bufferedStream);
+            } else if (lowerFileName.endsWith(".doc") || mimeType.equals("application/msword")) {
+                return parseDoc(bufferedStream);
+            } else if (lowerFileName.endsWith(".xlsx") || mimeType.equals("application/vnd.openxmlformats-officedocument.spreadsheetml.sheet")) {
+                return parseExcel(bufferedStream);
+            } else if (lowerFileName.endsWith(".xls") || mimeType.equals("application/vnd.ms-excel")) {
+                return parseExcel(bufferedStream);
+            } else if (lowerFileName.endsWith(".md") || lowerFileName.endsWith(".markdown") || mimeType.equals("text/markdown")) {
+                return parseMarkdown(bufferedStream);
+            } else if (lowerFileName.endsWith(".txt") || mimeType.startsWith("text/")) {
+                return parseText(bufferedStream);
+            } else {
+                // 使用 Tika 作为通用解析器
+                return tika.parseToString(bufferedStream);
+            }
+        } catch (Exception e) {
+            log.error("解析文档失败: {}", fileName, e);
+            throw new RuntimeException("解析文档失败: " + e.getMessage(), e);
+        }
+    }
+
+    /**
+     * 解析 PDF 文件
+     *
+     * @param inputStream PDF文件输入流
+     * @return 文本内容
+     */
+    public static String parsePdf(InputStream inputStream) {
+        try (PDDocument document = PDDocument.load(inputStream)) {
+            PDFTextStripper stripper = new PDFTextStripper();
+            // 设置编码，解决中文乱码问题
+            stripper.setSortByPosition(true);
+            String text = stripper.getText(document);
+            log.info("PDF解析成功，共 {} 页", document.getNumberOfPages());
+            return text;
+        } catch (Exception e) {
+            log.error("PDF解析失败", e);
+            throw new RuntimeException("PDF解析失败: " + e.getMessage(), e);
+        }
+    }
+
+    /**
+     * 解析 Word 2007+ 文件 (.docx)
+     *
+     * @param inputStream DOCX文件输入流
+     * @return 文本内容
+     */
+    public static String parseDocx(InputStream inputStream) {
+        try (XWPFDocument document = new XWPFDocument(inputStream)) {
+            StringBuilder text = new StringBuilder();
+            List<XWPFParagraph> paragraphs = document.getParagraphs();
+            for (XWPFParagraph paragraph : paragraphs) {
+                text.append(paragraph.getText()).append("\n");
+            }
+            log.info("DOCX解析成功，共 {} 段落", paragraphs.size());
+            return text.toString();
+        } catch (Exception e) {
+            log.error("DOCX解析失败", e);
+            throw new RuntimeException("DOCX解析失败: " + e.getMessage(), e);
+        }
+    }
+
+    /**
+     * 解析 Word 97-2003 文件 (.doc)
+     *
+     * @param inputStream DOC文件输入流
+     * @return 文本内容
+     */
+    public static String parseDoc(InputStream inputStream) {
+        try (HWPFDocument document = new HWPFDocument(inputStream);
+             WordExtractor extractor = new WordExtractor(document)) {
+            String text = extractor.getText();
+            log.info("DOC解析成功");
+            return text;
+        } catch (Exception e) {
+            log.error("DOC解析失败", e);
+            throw new RuntimeException("DOC解析失败: " + e.getMessage(), e);
+        }
+    }
+
+    /**
+     * 解析 Excel 文件 (.xlsx, .xls)
+     *
+     * @param inputStream Excel文件输入流
+     * @return 文本内容
+     */
+    public static String parseExcel(InputStream inputStream) {
+        try (Workbook workbook = WorkbookFactory.create(inputStream)) {
+            StringBuilder text = new StringBuilder();
+            int sheetCount = workbook.getNumberOfSheets();
+
+            for (int i = 0; i < sheetCount; i++) {
+                Sheet sheet = workbook.getSheetAt(i);
+                text.append("Sheet: ").append(sheet.getSheetName()).append("\n");
+
+                for (Row row : sheet) {
+                    StringBuilder rowText = new StringBuilder();
+                    for (Cell cell : row) {
+                        String cellValue = getCellValueAsString(cell);
+                        if (!cellValue.isEmpty()) {
+                            rowText.append(cellValue).append("\t");
+                        }
+                    }
+                    if (rowText.length() > 0) {
+                        text.append(rowText.toString().trim()).append("\n");
+                    }
+                }
+                text.append("\n");
+            }
+
+            log.info("Excel解析成功，共 {} 个Sheet", sheetCount);
+            return text.toString();
+        } catch (Exception e) {
+            log.error("Excel解析失败", e);
+            throw new RuntimeException("Excel解析失败: " + e.getMessage(), e);
+        }
+    }
+
+    /**
+     * 获取单元格的字符串值
+     */
+    private static String getCellValueAsString(Cell cell) {
+        if (cell == null) {
+            return "";
+        }
+
+        switch (cell.getCellType()) {
+            case STRING:
+                return cell.getStringCellValue();
+            case NUMERIC:
+                if (DateUtil.isCellDateFormatted(cell)) {
+                    return cell.getDateCellValue().toString();
+                }
+                return String.valueOf(cell.getNumericCellValue());
+            case BOOLEAN:
+                return String.valueOf(cell.getBooleanCellValue());
+            case FORMULA:
+                return cell.getCellFormula();
+            default:
+                return "";
+        }
+    }
+
+    /**
+     * 解析 Markdown 文件
+     *
+     * @param inputStream Markdown文件输入流
+     * @return 文本内容
+     */
+    public static String parseMarkdown(InputStream inputStream) {
+        try {
+            String content = new String(inputStream.readAllBytes(), java.nio.charset.StandardCharsets.UTF_8);
+
+            // 配置 Flexmark
+            MutableDataSet options = new MutableDataSet();
+            Parser parser = Parser.builder(options).build();
+            HtmlRenderer renderer = HtmlRenderer.builder(options).build();
+
+            // 解析 Markdown 为 HTML（可选，如果需要纯文本可以直接返回原始内容）
+            com.vladsch.flexmark.util.ast.Node document = parser.parse(content);
+            String html = renderer.render(document);
+
+            log.info("Markdown解析成功");
+            // 返回原始 Markdown 文本，如需 HTML 可返回 html 变量
+            return content;
+        } catch (Exception e) {
+            log.error("Markdown解析失败", e);
+            throw new RuntimeException("Markdown解析失败: " + e.getMessage(), e);
+        }
+    }
+
+    /**
+     * 解析纯文本文件
+     *
+     * @param inputStream 文本文件输入流
+     * @return 文本内容
+     */
+    public static String parseText(InputStream inputStream) {
+        try {
+            String content = new String(inputStream.readAllBytes(), java.nio.charset.StandardCharsets.UTF_8);
+            log.info("文本文件解析成功");
+            return content;
+        } catch (Exception e) {
+            log.error("文本文件解析失败", e);
+            throw new RuntimeException("文本文件解析失败: " + e.getMessage(), e);
+        }
+    }
+
+    /**
+     * 检测文件类型
+     *
+     * @param inputStream 文件输入流
+     * @param fileName    文件名
+     * @return MIME类型
+     */
+    public static String detectMimeType(InputStream inputStream, String fileName) {
+        try {
+            return tika.detect(inputStream, fileName);
+        } catch (Exception e) {
+            log.error("检测文件类型失败", e);
+            return "application/octet-stream";
+        }
+    }
+}
diff --git a/src/main/java/cn/yinlihupo/controller/project/ProjectController.java b/src/main/java/cn/yinlihupo/controller/project/ProjectController.java
index 3090541..fa7e543 100644
--- a/src/main/java/cn/yinlihupo/controller/project/ProjectController.java
+++ b/src/main/java/cn/yinlihupo/controller/project/ProjectController.java
@@ -23,29 +23,6 @@ public class ProjectController {
     private final ProjectService projectService;
     private final OssService ossService;
 
-    /**
-     * 根据文本内容生成项目初始化数据
-     *
-     * @param request 包含项目资料内容的请求
-     * @return 项目初始化结构化数据
-     */
-    @PostMapping("/from-content")
-    public Result<ProjectInitResult> generateFromContent(@RequestBody ProjectInitRequest request) {
-        log.info("收到项目初始化请求（文本内容）");
-
-        if (request.getContent() == null || request.getContent().trim().isEmpty()) {
-            return Result.error("项目资料内容不能为空");
-        }
-
-        try {
-            ProjectInitResult result = projectService.generateProjectFromContent(request.getContent());
-            return Result.success("项目初始化成功", result);
-        } catch (Exception e) {
-            log.error("项目初始化失败: {}", e.getMessage(), e);
-            return Result.error("项目初始化失败: " + e.getMessage());
-        }
-    }
-
     /**
      * 上传文件并生成项目初始化数据
      *
diff --git a/src/main/java/cn/yinlihupo/service/oss/impl/OssServiceImpl.java b/src/main/java/cn/yinlihupo/service/oss/impl/OssServiceImpl.java
index f8e25f8..8c36ef8 100644
--- a/src/main/java/cn/yinlihupo/service/oss/impl/OssServiceImpl.java
+++ b/src/main/java/cn/yinlihupo/service/oss/impl/OssServiceImpl.java
@@ -1,6 +1,7 @@
 package cn.yinlihupo.service.oss.impl;
 
 import cn.yinlihupo.common.config.MinioConfig;
+import cn.yinlihupo.common.util.DocumentParserUtil;
 import cn.yinlihupo.service.oss.OssService;
 import io.minio.*;
 import io.minio.errors.*;
@@ -61,22 +62,43 @@ public class OssServiceImpl implements OssService {
 
     @Override
     public String readFileAsString(String fileUrl) {
-        try (InputStream inputStream = getFileInputStream(fileUrl);
-             ByteArrayOutputStream outputStream = new ByteArrayOutputStream()) {
-
-            byte[] buffer = new byte[1024];
-            int bytesRead;
-            while ((bytesRead = inputStream.read(buffer)) != -1) {
-                outputStream.write(buffer, 0, bytesRead);
-            }
-
-            return outputStream.toString(StandardCharsets.UTF_8);
+        try (InputStream inputStream = getFileInputStream(fileUrl)) {
+            // 从 URL 中提取文件名
+            String fileName = extractFileNameFromUrl(fileUrl);
+            // 使用文档解析工具类解析文件内容
+            return DocumentParserUtil.parse(inputStream, fileName);
         } catch (Exception e) {
             log.error("读取文件内容失败: {}", e.getMessage(), e);
             throw new RuntimeException("读取文件内容失败: " + e.getMessage(), e);
         }
     }
 
+    /**
+     * 从 URL 中提取文件名
+     *
+     * @param fileUrl 文件 URL
+     * @return 文件名
+     */
+    private String extractFileNameFromUrl(String fileUrl) {
+        try {
+            URL url = new URL(fileUrl);
+            String path = url.getPath();
+            // 去掉开头的 /
+            if (path.startsWith("/")) {
+                path = path.substring(1);
+            }
+            // 获取最后一个 / 后面的文件名
+            int lastSlashIndex = path.lastIndexOf('/');
+            if (lastSlashIndex >= 0) {
+                return path.substring(lastSlashIndex + 1);
+            }
+            return path;
+        } catch (Exception e) {
+            log.warn("从 URL 提取文件名失败: {}", fileUrl);
+            return "unknown";
+        }
+    }
+
     @Override
     public InputStream getFileInputStream(String fileUrl) {
         try {
diff --git a/src/main/java/cn/yinlihupo/service/project/impl/ProjectServiceImpl.java b/src/main/java/cn/yinlihupo/service/project/impl/ProjectServiceImpl.java
index db31d05..e6b48aa 100644
--- a/src/main/java/cn/yinlihupo/service/project/impl/ProjectServiceImpl.java
+++ b/src/main/java/cn/yinlihupo/service/project/impl/ProjectServiceImpl.java
@@ -147,8 +147,10 @@ public class ProjectServiceImpl implements ProjectService {
     public ProjectInitResult generateProjectFromContent(String content) {
         log.info("开始根据内容生成项目初始化数据");
 
-        PromptTemplate promptTemplate = new PromptTemplate(USER_PROMPT_TEMPLATE);
-        String userPrompt = promptTemplate.createMessage(java.util.Map.of("content", content)).toString();
+        // 构建用户提示词，直接将内容嵌入
+        String userPrompt = "请根据以下项目资料，生成完整的项目初始化结构化数据：\n\n" +
+                content + "\n\n" +
+                "请严格按照系统提示词中的JSON格式输出，确保所有字段都包含合理的值。";
 
         return chatClient.prompt()
                 .system(PROJECT_INIT_SYSTEM_PROMPT)
diff --git a/src/main/resources/application-dev.yaml b/src/main/resources/application-dev.yaml
index f396cb1..5e09858 100644
--- a/src/main/resources/application-dev.yaml
+++ b/src/main/resources/application-dev.yaml
@@ -37,12 +37,11 @@ spring:
       base-url: https://sg1.proxy.yinlihupo.cc/proxy/https://openrouter.ai/api/v1
       chat:
         options:
-          model: gpt-4o
-          temperature: 0.3
+          model: Qwen3
 
 # MinIO 对象存储配置
 minio:
-  endpoint: 10.200.8.25:9000
+  endpoint: http://10.200.8.25:9000
   access-key: minioadmin
   secret-key: minioadmin
   bucket-name: ylhp-files