@saber2pr/ai-agent 0.0.44 → 0.0.46

package/lib/core/agent-graph.js

@@ -426,12 +426,9 @@ class McpGraphAgent {
 3. **禁止空操作**：如果你认为任务已完成或不需要调用工具，请不要输出任何 Action 结构。严禁使用 "None"、"null" 或空字符串作为工具名称。
 
 # 🎯 核心指令
-1. **任务终结判定**：当你已经读取了用户要求的文件、回答了问题或完成了审计目标时，必须立即提供最终回复。
+1. **任务终结判定**：当你已经读取了用户要求的文件、回答了问题或完成了代码编写时，必须立即提供最终回复。
 2. **回复格式**：任务完成时，请以 "Final Answer:" 开头进行总结，此时不再调用任何工具。
 
-# 📝 审计任务专项
-1. 避免在同一个文件上陷入无限循环尝试。
-2. 优先通过 \`list_directory\` 了解全局，再深入具体文件。
 {extraPrompt}`;
         // 2. 核心逻辑：处理消息上下文
         let inputMessages;

package/lib/tools/filesystem/index.js

@@ -4,7 +4,6 @@ var __importDefault = (this && this.__importDefault) || function (mod) {
 };
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.getFilesystemTools = void 0;
-/* eslint-disable no-continue */
 const promises_1 = __importDefault(require("fs/promises"));
 const minimatch_1 = require("minimatch");
 const path_1 = __importDefault(require("path"));
@@ -27,25 +26,9 @@ const WriteFileArgsSchema = zod_1.z.object({
     path: zod_1.z.string(),
     content: zod_1.z.string(),
 });
-const EditOperation = zod_1.z.object({
-    oldText: zod_1.z.string().describe('Text to search for - must match exactly'),
-    newText: zod_1.z.string().describe('Text to replace with'),
-});
-const EditFileArgsSchema = zod_1.z.object({
-    path: zod_1.z.string(),
-    edits: zod_1.z.array(EditOperation),
-    dryRun: zod_1.z
-        .boolean()
-        .optional()
-        .default(false)
-        .describe('Preview changes using git-style diff format'),
-});
 const CreateDirectoryArgsSchema = zod_1.z.object({
     path: zod_1.z.string(),
 });
-const ListDirectoryArgsSchema = zod_1.z.object({
-    path: zod_1.z.string(),
-});
 const ListDirectoryWithSizesArgsSchema = zod_1.z.object({
     path: zod_1.z.string(),
     sortBy: zod_1.z
@@ -57,6 +40,7 @@ const ListDirectoryWithSizesArgsSchema = zod_1.z.object({
 const DirectoryTreeArgsSchema = zod_1.z.object({
     path: zod_1.z.string(),
     excludePatterns: zod_1.z.array(zod_1.z.string()).optional().default([]),
+    depth: zod_1.z.number().optional().default(2).describe('递归深度，默认 2 层。增加深度会消耗更多 Token'),
 });
 const MoveFileArgsSchema = zod_1.z.object({
     source: zod_1.z.string(),
@@ -71,9 +55,19 @@ const GetFileInfoArgsSchema = zod_1.z.object({
     path: zod_1.z.string(),
 });
 const GrepSearchArgsSchema = zod_1.z.object({
-    path: zod_1.z.string().describe('The starting directory path for searching content'),
-    query: zod_1.z.string().describe('The text keyword to search for'),
-    includePattern: zod_1.z.string().optional().default('**/*').describe('Only search in files matching this pattern, for example "**/*.ts"'),
+    path: zod_1.z.string().describe('搜索的起始目录路径'),
+    query: zod_1.z.string().describe('要搜索的文本关键字'),
+    includePattern: zod_1.z.string().optional().default('**/*').describe('匹配模式，例如 "**/*.ts"'),
+    maxFiles: zod_1.z.number().optional().default(100).describe('最大扫描文件数，防止大型项目超时'),
+});
+const PatchEditArgsSchema = zod_1.z.object({
+    path: zod_1.z.string().describe('文件路径'),
+    patches: zod_1.z.array(zod_1.z.object({
+        startLine: zod_1.z.number().describe('起始行号（包含）'),
+        endLine: zod_1.z.number().describe('结束行号（包含）'),
+        replacement: zod_1.z.string().describe('要插入的新代码内容'),
+        originalSnippet: zod_1.z.string().optional().describe('可选：该行范围内的原始代码片段，用于二次校验防止行号偏移'),
+    })).describe('补丁列表。注意：若有多个补丁，建议从文件尾部向头部执行，或确保行号不重叠'),
 });
 const getFilesystemTools = (targetDir) => {
     (0, lib_1.setAllowedDirectories)([targetDir]);
@@ -97,44 +91,48 @@ const getFilesystemTools = (targetDir) => {
     };
     const readTextFileTool = (0, createTool_1.createTool)({
         name: 'read_text_file',
-        description: 'Read the complete contents of a file from the file system as text. ' +
-            'Handles various text encodings and provides detailed error messages ' +
-            'if the file cannot be read. Use this tool when you need to examine ' +
-            "the contents of a single file. Use the 'head' parameter to read only " +
-            "the first N lines of a file, or the 'tail' parameter to read only " +
-            'the last N lines of a file. Operates on the file as text regardless of extension.',
+        description: '读取文件全文。若超过100行则禁止使用，必须改用 read_file_range。支持 head/tail 参数。',
         parameters: ReadTextFileArgsSchema,
         handler: readTextFileHandler,
     });
     const readMultipleFilesTool = (0, createTool_1.createTool)({
         name: 'read_multiple_files',
-        description: 'Read the contents of multiple files simultaneously. This is more ' +
-            'efficient than reading files one by one when you need to analyze ' +
-            "or compare multiple files. Each file's content is returned with its " +
-            "path as a reference. Failed reads for individual files won't stop " +
-            'the entire operation. Only works within allowed directories.',
+        description: '同时读取多个文件的内容。当你需要对比多个文件或分析跨文件关联时使用。' +
+            '注意：为了防止 Token 溢出，本工具一次最多读取 10 个文件，且每个文件仅展示前 6000 字符。' +
+            '若需查看完整大文件或特定逻辑，请改用 read_file_range。',
         parameters: ReadMultipleFilesArgsSchema,
         handler: async (args) => {
-            const results = await Promise.all(args.paths.map(async (filePath) => {
+            // 保护 1：文件数量限制 (防止 AI 一次传入几十个文件)
+            const MAX_FILES = 10;
+            const pathsToRead = args.paths.slice(0, MAX_FILES);
+            const isTruncatedByCount = args.paths.length > MAX_FILES;
+            // 保护 2：单文件字符数限制 (防止读入超大型二进制或日志文件)
+            const MAX_CHARS_PER_FILE = 6000;
+            const results = await Promise.all(pathsToRead.map(async (filePath) => {
                 try {
+                    // 沿用你现有的路径验证逻辑
                     const validPath = await (0, lib_1.validatePath)(targetDir, filePath);
                     const content = await (0, lib_1.readFileContent)(validPath);
+                    if (content.length > MAX_CHARS_PER_FILE) {
+                        return `${filePath} (内容已截断):\n${content.substring(0, MAX_CHARS_PER_FILE)}\n\n[... 内容过长，仅展示前 ${MAX_CHARS_PER_FILE} 字符。若需查看后续内容，请使用 read_file_range 指定行号读取 ...]`;
+                    }
                     return `${filePath}:\n${content}\n`;
                 }
                 catch (error) {
                     const errorMessage = error instanceof Error ? error.message : String(error);
-                    return `${filePath}: Error - ${errorMessage}`;
+                    return `${filePath}: 读取失败 - ${errorMessage}`;
                 }
             }));
-            const text = results.join('\n---\n');
+            let text = results.join('\n---\n');
+            if (isTruncatedByCount) {
+                text += `\n\n⚠️ 注意：一次请求最多处理 ${MAX_FILES} 个文件。剩余 ${args.paths.length - MAX_FILES} 个文件未读取，请分批请求。`;
+            }
             return text;
         },
     });
     const writeFileTool = (0, createTool_1.createTool)({
         name: 'write_file',
-        description: 'Create a new file or completely overwrite an existing file with new content. ' +
-            'Use with caution as it will overwrite existing files without warning. ' +
-            'Handles text content with proper encoding. Only works within allowed directories.',
+        description: '仅用于创建新文件。严禁用于修改现有源代码。',
         parameters: WriteFileArgsSchema,
         handler: async (args) => {
             const validPath = await (0, lib_1.validatePath)(targetDir, args.path);
@@ -143,18 +141,6 @@ const getFilesystemTools = (targetDir) => {
             return text;
         },
     });
-    const editFileTool = (0, createTool_1.createTool)({
-        name: 'edit_file',
-        description: 'Make line-based edits to a text file. Each edit replaces exact line sequences ' +
-            'with new content. Returns a git-style diff showing the changes made. ' +
-            'Only works within allowed directories.',
-        parameters: EditFileArgsSchema,
-        handler: async (args) => {
-            const validPath = await (0, lib_1.validatePath)(targetDir, args.path);
-            const result = await (0, lib_1.applyFileEdits)(validPath, args.edits, args.dryRun);
-            return result;
-        },
-    });
     const createDirectoryTool = (0, createTool_1.createTool)({
         name: 'create_directory',
         description: 'Create a new directory or ensure a directory exists. Can create multiple ' +
@@ -169,24 +155,8 @@ const getFilesystemTools = (targetDir) => {
             return text;
         },
     });
-    const listDirectoryTool = (0, createTool_1.createTool)({
-        name: 'list_directory',
-        description: 'Get a detailed listing of all files and directories in a specified path. ' +
-            'Results clearly distinguish between files and directories with [FILE] and [DIR] ' +
-            'prefixes. This tool is essential for understanding directory structure and ' +
-            'finding specific files within a directory. Only works within allowed directories.',
-        parameters: ListDirectoryArgsSchema,
-        handler: async (args) => {
-            const validPath = await (0, lib_1.validatePath)(targetDir, args.path);
-            const entries = await promises_1.default.readdir(validPath, { withFileTypes: true });
-            const formatted = entries
-                .map(entry => `${entry.isDirectory() ? '[DIR]' : '[FILE]'} ${entry.name}`)
-                .join('\n');
-            return formatted;
-        },
-    });
     const listDirectoryWithSizesTool = (0, createTool_1.createTool)({
-        name: 'list_directory_with_sizes',
+        name: 'list_directory',
         description: 'Get a detailed listing of all files and directories in a specified path, including sizes. ' +
             'Results clearly distinguish between files and directories with [FILE] and [DIR] ' +
             'prefixes. This tool is useful for understanding directory structure and ' +
@@ -241,46 +211,36 @@ const getFilesystemTools = (targetDir) => {
     });
     const directoryTreeTool = (0, createTool_1.createTool)({
         name: 'directory_tree',
-        description: 'Get a recursive tree view of files and directories as a JSON structure. ' +
-            "Each entry includes 'name', 'type' (file/directory), and 'children' for directories. " +
-            'Files have no children array, while directories always have a children array (which may be empty). ' +
-            'The output is formatted with 2-space indentation for readability. Only works within allowed directories.',
+        description: '获取目录的递归树状 JSON 结构。' +
+            '默认仅展示 2 层深度以节省 Token。如果需要看更深层级，请调大 depth 参数。',
         parameters: DirectoryTreeArgsSchema,
         handler: async (args) => {
             const rootPath = args.path;
-            async function buildTree(currentPath, excludePatterns = []) {
+            async function buildTree(currentPath, currentDepth, maxDepth, excludePatterns = []) {
+                if (currentDepth > maxDepth)
+                    return []; // 深度限制
                 const validPath = await (0, lib_1.validatePath)(targetDir, currentPath);
                 const entries = await promises_1.default.readdir(validPath, { withFileTypes: true });
                 const result = [];
                 for (const entry of entries) {
                     const relativePath = path_1.default.relative(rootPath, path_1.default.join(currentPath, entry.name));
-                    const shouldExclude = excludePatterns.some(pattern => {
-                        if (pattern.includes('*')) {
-                            return (0, minimatch_1.minimatch)(relativePath, pattern, { dot: true });
-                        }
-                        // For files: match exact name or as part of path
-                        // For directories: match as directory path
-                        return ((0, minimatch_1.minimatch)(relativePath, pattern, { dot: true }) ||
-                            (0, minimatch_1.minimatch)(relativePath, `**/${pattern}`, { dot: true }) ||
-                            (0, minimatch_1.minimatch)(relativePath, `**/${pattern}/**`, { dot: true }));
-                    });
+                    const shouldExclude = excludePatterns.some(pattern => (0, minimatch_1.minimatch)(relativePath, pattern, { dot: true }));
                     if (shouldExclude)
                         continue;
                     const entryData = {
                         name: entry.name,
                         type: entry.isDirectory() ? 'directory' : 'file',
                     };
-                    if (entry.isDirectory()) {
+                    if (entry.isDirectory() && currentDepth < maxDepth) {
                         const subPath = path_1.default.join(currentPath, entry.name);
-                        entryData.children = await buildTree(subPath, excludePatterns);
+                        entryData.children = await buildTree(subPath, currentDepth + 1, maxDepth, excludePatterns);
                     }
                     result.push(entryData);
                 }
                 return result;
             }
-            const treeData = await buildTree(rootPath, args.excludePatterns);
-            const text = JSON.stringify(treeData, null, 2);
-            return text;
+            const treeData = await buildTree(rootPath, 1, args.depth, args.excludePatterns);
+            return JSON.stringify(treeData, null, 2);
         },
     });
     const moveFileTool = (0, createTool_1.createTool)({
@@ -301,7 +261,8 @@ const getFilesystemTools = (targetDir) => {
     const searchFilesTool = (0, createTool_1.createTool)({
         name: 'search_files',
         description: 'Search for files matching a specific pattern in a specified path. ' +
-            'Returns a list of files that match the pattern. Only works within allowed directories.',
+            'Returns a list of files that match the pattern. Only works within allowed directories.' +
+            'Used only for filename search',
         parameters: SearchFilesArgsSchema,
         handler: async (args) => {
             const validPath = await (0, lib_1.validatePath)(targetDir, args.path);
@@ -314,68 +275,152 @@ const getFilesystemTools = (targetDir) => {
     });
     const getFileInfoTool = (0, createTool_1.createTool)({
         name: 'get_file_info',
-        description: 'Get detailed information about a file, including its size, last modified time, and type. ' +
-            'Only works within allowed directories.',
+        description: '查看文件元数据（大小、行数、修改时间）。读取大文件前务必先调用此工具。',
         parameters: GetFileInfoArgsSchema,
         handler: async (args) => {
             const validPath = await (0, lib_1.validatePath)(targetDir, args.path);
-            const info = await (0, lib_1.getFileStats)(validPath);
-            const text = Object.entries(info)
+            const stats = await promises_1.default.stat(validPath);
+            // 计算行数：读取内容并按换行符分割
+            // 注意：对于极大的文件，这种方式可能稍慢，但对普通源代码文件非常有效
+            const content = await promises_1.default.readFile(validPath, 'utf-8');
+            const lineCount = content.split('\n').length;
+            const info = {
+                size: `${(stats.size / 1024).toFixed(2)} KB`,
+                lineCount: lineCount, // 新增行号字段
+                mtime: stats.mtime.toLocaleString(),
+                type: path_1.default.extname(validPath) || 'unknown'
+            };
+            return Object.entries(info)
                 .map(([key, value]) => `${key}: ${value}`)
                 .join('\n');
-            return text;
         },
     });
     const grepSearchTool = (0, createTool_1.createTool)({
         name: 'grep_search',
-        description: 'Search for a specific keyword in the content of files in a specified directory. ' +
-            'This tool will recursively read files and return the paths of all files containing the keyword. ' +
-            'Useful for finding variable definitions, function calls, or specific text.',
+        description: '在指定目录的文件内容中搜索关键字。' +
+            '该工具会返回包含关键字的文件路径及匹配行的预览。' +
+            '请尽量通过 includePattern 缩小搜索范围。',
         parameters: GrepSearchArgsSchema,
         handler: async (args) => {
             const startPath = await (0, lib_1.validatePath)(targetDir, args.path);
-            // 1. 利用你现有的工具先筛选出文件列表
-            const allFiles = await (0, lib_1.searchFilesWithValidation)(targetDir, startPath, args.includePattern, [targetDir], { excludePatterns: ['node_modules', 'dist', '.git'] });
+            const allFiles = await (0, lib_1.searchFilesWithValidation)(targetDir, startPath, args.includePattern, [targetDir], { excludePatterns: ['node_modules', 'dist', '.git', 'build'] });
+            // 限制扫描文件数，防止爆炸
+            const filesToScan = allFiles.slice(0, args.maxFiles);
             const matches = [];
-            // 2. 并发读取文件内容进行关键词匹配 (限制并发数为 10，防止内存或句柄爆炸)
             const concurrencyLimit = 10;
-            for (let i = 0; i < allFiles.length; i += concurrencyLimit) {
-                const chunk = allFiles.slice(i, i + concurrencyLimit);
+            for (let i = 0; i < filesToScan.length; i += concurrencyLimit) {
+                const chunk = filesToScan.slice(i, i + concurrencyLimit);
                 await Promise.all(chunk.map(async (filePath) => {
                     try {
-                        // 注意：这里确保只搜索文件，不搜索目录
                         const stats = await promises_1.default.stat(filePath);
                         if (!stats.isFile())
                             return;
                         const content = await (0, lib_1.readFileContent)(filePath);
                         if (content.includes(args.query)) {
-                            // 返回相对路径，方便查看
-                            matches.push(path_1.default.relative(targetDir, filePath));
+                            const relativePath = path_1.default.relative(targetDir, filePath);
+                            // 找到匹配的那一行（预览用）
+                            const lines = content.split('\n');
+                            const matchLineIndex = lines.findIndex(l => l.includes(args.query));
+                            matches.push(`${relativePath} (Line ${matchLineIndex + 1}: "${lines[matchLineIndex].trim().substring(0, 100)}")`);
                         }
                     }
-                    catch {
-                        // 忽略读取失败的文件（如二进制文件或无权限文件）
-                    }
+                    catch { /* 忽略错误文件 */ }
                 }));
             }
-            return matches.length > 0
-                ? `Found keyword "${args.query}" in the following files:\n${matches.join('\n')}`
-                : `No content containing "${args.query}" found in the specified range.`;
+            let response = matches.length > 0
+                ? `找到关键词 "${args.query}" 的位置如下：\n${matches.join('\n')}`
+                : `未找到包含 "${args.query}" 的内容。`;
+            if (allFiles.length > args.maxFiles) {
+                response += `\n\n注意：搜索已达到限制，仅扫描了前 ${args.maxFiles} 个文件。若未找到结果，请提供更精确的 path 或 includePattern。`;
+            }
+            return response;
+        },
+    });
+    const readFileRangeTool = (0, createTool_1.createTool)({
+        name: 'read_file_range',
+        description: '精准读取指定行范围（包含行号前缀）。修改代码前或根据报错定位时必用。',
+        parameters: zod_1.z.object({
+            path: zod_1.z.string().describe('相对于目标目录的文件路径'),
+            startLine: zod_1.z.number().describe('起始行号（从 1 开始计）'),
+            endLine: zod_1.z.number().describe('结束行号'),
+        }),
+        handler: async (args) => {
+            // 1. 验证路径安全（沿用你代码中的 validatePath 逻辑）
+            const validPath = await (0, lib_1.validatePath)(targetDir, args.path);
+            try {
+                const content = await promises_1.default.readFile(validPath, 'utf-8');
+                const lines = content.split('\n');
+                const totalLines = lines.length;
+                // 2. 边界保护：确保行号不越界
+                const start = Math.max(1, args.startLine);
+                const end = Math.min(totalLines, args.endLine);
+                if (start > totalLines) {
+                    return `错误：文件仅有 ${totalLines} 行，起始行号 ${start} 超出范围。`;
+                }
+                if (start > end) {
+                    return `错误：起始行号 ${start} 不能大于结束行号 ${end}。`;
+                }
+                // 3. 截取并添加行号索引（核心：增强 AI 的位置感）
+                const selectedLines = lines.slice(start - 1, end);
+                const formattedContent = selectedLines
+                    .map((line, index) => `${start + index}| ${line}`)
+                    .join('\n');
+                return `[文件: ${args.path} | 第 ${start} 至 ${end} 行 / 共 ${totalLines} 行]\n${formattedContent}`;
+            }
+            catch (error) {
+                return `读取文件范围失败: ${error.message}`;
+            }
+        },
+    });
+    const editFileTool = (0, createTool_1.createTool)({
+        name: 'edit_file',
+        description: '基于行号范围替换代码。修改逻辑的唯一工具。调用前须通过 read_file_range 获取最新行号。支持删除(空内容)或单行替换。',
+        parameters: PatchEditArgsSchema,
+        handler: async (args) => {
+            const validPath = await (0, lib_1.validatePath)(targetDir, args.path);
+            try {
+                const content = await promises_1.default.readFile(validPath, 'utf-8');
+                let lines = content.split('\n');
+                // 按起始行号从大到小排序，这样修改前面的行不会影响后面待修改行的索引
+                const sortedPatches = [...args.patches].sort((a, b) => b.startLine - a.startLine);
+                for (const patch of sortedPatches) {
+                    // 校验行号合法性
+                    if (patch.startLine < 1 || patch.endLine > lines.length || patch.startLine > patch.endLine) {
+                        return `错误：行号范围 ${patch.startLine}-${patch.endLine} 超出文件实际范围 (1-${lines.length})`;
+                    }
+                    // 可选：二次校验（防止 AI 记忆了错误的行号）
+                    if (patch.originalSnippet) {
+                        const currentText = lines.slice(patch.startLine - 1, patch.endLine).join('\n');
+                        // 模糊对比，如果差异太大则报错
+                        if (!currentText.includes(patch.originalSnippet.trim()) && currentText.trim().length > 0) {
+                            return `警告：第 ${patch.startLine} 行的内容已发生变动，与你预想的代码不符。请重新读取文件获取最新行号。`;
+                        }
+                    }
+                    // 执行替换：splice(开始索引, 删除数量, 替换内容)
+                    // 索引需要减 1
+                    lines.splice(patch.startLine - 1, (patch.endLine - patch.startLine) + 1, patch.replacement);
+                }
+                await promises_1.default.writeFile(validPath, lines.join('\n'), 'utf-8');
+                return `成功通过行号更新了 ${args.path} 的 ${args.patches.length} 处代码。`;
+            }
+            catch (error) {
+                return `Patch 失败: ${error.message}`;
+            }
         },
     });
     return [
+        readFileRangeTool,
+        editFileTool,
+        directoryTreeTool,
+        listDirectoryWithSizesTool,
+        grepSearchTool,
+        getFileInfoTool,
         readTextFileTool,
         readMultipleFilesTool,
+        searchFilesTool,
         writeFileTool,
-        editFileTool,
         createDirectoryTool,
-        listDirectoryTool,
-        listDirectoryWithSizesTool,
-        directoryTreeTool,
         moveFileTool,
-        searchFilesTool,
-        getFileInfoTool,
-        grepSearchTool,
     ];
 };
 exports.getFilesystemTools = getFilesystemTools;

package/lib/tools/ts-lsp/index.js

@@ -8,8 +8,19 @@ const getTsLspTools = (targetDir) => {
     const engine = new ts_context_mcp_1.PromptEngine(targetDir);
     return [
         (0, createTool_1.createTool)({
-            name: "get_repo_map",
-            description: "获取项目全局文件结构及导出清单，用于快速定位代码",
+            name: 'get_method_body',
+            description: '【仅限TS/JS】通过方法名提取代码块。比行号读取更抗干扰，参考逻辑时首选。',
+            parameters: zod_1.z.object({
+                filePath: zod_1.z.string().describe('文件相对路径'),
+                methodName: zod_1.z.string().describe('方法名或函数名'),
+            }),
+            handler: async ({ filePath, methodName }) => {
+                return engine.getMethodImplementation(filePath, methodName);
+            },
+        }),
+        (0, createTool_1.createTool)({
+            name: 'get_repo_map',
+            description: '获取项目全局文件结构及导出清单，用于快速定位代码',
             parameters: zod_1.z.object({}),
             handler: async () => {
                 engine.refresh();
@@ -17,18 +28,18 @@ const getTsLspTools = (targetDir) => {
             },
         }),
         (0, createTool_1.createTool)({
-            name: "analyze_deps",
-            description: "分析指定文件的依赖关系，支持 tsconfig 路径别名解析",
-            parameters: zod_1.z.object({ filePath: zod_1.z.string().describe("文件相对路径") }),
+            name: 'analyze_deps',
+            description: '分析指定文件的依赖关系，支持 tsconfig 路径别名解析',
+            parameters: zod_1.z.object({ filePath: zod_1.z.string().describe('文件相对路径') }),
             handler: async ({ filePath }) => engine.getDeps(filePath),
         }),
         (0, createTool_1.createTool)({
-            name: "read_skeleton",
-            description: "提取文件的结构定义（接口、类、方法签名），忽略具体实现以节省 Token",
-            parameters: zod_1.z.object({ filePath: zod_1.z.string().describe("文件相对路径") }),
+            name: 'read_skeleton',
+            description: '提取文件的结构定义（接口、类、方法签名），忽略具体实现以节省 Token',
+            parameters: zod_1.z.object({ filePath: zod_1.z.string().describe('文件相对路径') }),
             handler: async (args) => {
                 const pathArg = args === null || args === void 0 ? void 0 : args.filePath;
-                if (typeof pathArg !== "string" || pathArg.trim() === "") {
+                if (typeof pathArg !== 'string' || pathArg.trim() === '') {
                     return `Error: 参数 'filePath' 无效。收到的是: ${JSON.stringify(pathArg)}`;
                 }
                 try {
@@ -41,14 +52,6 @@ const getTsLspTools = (targetDir) => {
                 }
             },
         }),
-        (0, createTool_1.createTool)({
-            name: "get_method_body",
-            description: "获取指定文件内某个方法或函数的完整实现代码",
-            parameters: zod_1.z.object({ filePath: zod_1.z.string().describe("文件相对路径"), methodName: zod_1.z.string().describe("方法名或函数名") }),
-            handler: async ({ filePath, methodName }) => {
-                return engine.getMethodImplementation(filePath, methodName);
-            },
-        }),
     ];
 };
 exports.getTsLspTools = getTsLspTools;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@saber2pr/ai-agent",
-  "version": "0.0.44",
+  "version": "0.0.46",
   "description": "AI Assistant CLI.",
   "author": "saber2pr",
   "license": "ISC",