Add .git-ai/include.txt to force-index gitignored directories

README.md

@@ -286,6 +286,79 @@ That's it! 3 steps to get started, immediately begin deep understanding of your
 
 ---
 
+## ⚙️ Configuration
+
+### File Filtering
+
+git-ai respects your project's ignore files to control which files are indexed:
+
+#### `.gitignore` - Standard Git Ignore
+
+Files matching patterns in `.gitignore` are excluded from indexing by default.
+
+#### `.aiignore` - AI-Specific Exclusions (Highest Priority)
+
+Create a `.aiignore` file in your repository root to exclude specific files from indexing that should be ignored by git-ai but not necessarily by Git:
+
+```bash
+# Example .aiignore
+test-fixtures/**
+*.generated.ts
+docs/api-reference/**
+```
+
+#### `.git-ai/include.txt` - Force Include (Overrides `.gitignore`)
+
+Sometimes you need to index generated code or files that are in `.gitignore` but important for code understanding. Create `.git-ai/include.txt` to force-index specific patterns:
+
+```bash
+# Example .git-ai/include.txt
+# Include generated API clients
+generated/api/**
+
+# Include specific build artifacts that contain important types
+dist/types/**
+
+# Include code from specific ignored directories
+vendor/important-lib/**
+```
+
+**Priority Order (Highest to Lowest):**
+1. `.aiignore` - Explicit exclusions always win
+2. `.git-ai/include.txt` - Force-include patterns override `.gitignore`
+3. `.gitignore` - Standard Git ignore patterns
+
+**Supported Pattern Syntax:**
+- `**` - Match any number of directories
+- `*` - Match any characters within a directory
+- `directory/` - Match entire directory (automatically converts to `directory/**`)
+- `file.ts` - Match specific file
+- Lines starting with `#` are comments
+
+**Example Configuration:**
+
+```bash
+# .gitignore
+dist/
+generated/
+*.log
+
+# .git-ai/include.txt
+generated/api/**
+generated/types/**
+
+# .aiignore (overrides everything)
+generated/test-data/**
+```
+
+With this configuration:
+- ✅ `generated/api/client.ts` - Indexed (included via include.txt)
+- ✅ `generated/types/models.ts` - Indexed (included via include.txt)
+- ❌ `generated/test-data/mock.ts` - Not indexed (.aiignore takes priority)
+- ❌ `dist/bundle.js` - Not indexed (.gitignore, not in include.txt)
+
+---
+
 ## 🛠️ Troubleshooting
 
 ### Windows Installation Issues

README.zh-CN.md

@@ -283,6 +283,79 @@ git-ai ai graph callers authenticateUser
 
 ---
 
+## ⚙️ 配置
+
+### 文件过滤
+
+git-ai 遵循项目的忽略文件配置来控制哪些文件会被索引：
+
+#### `.gitignore` - 标准 Git 忽略规则
+
+默认情况下，匹配 `.gitignore` 中模式的文件会被排除在索引之外。
+
+#### `.aiignore` - AI 专属排除规则（最高优先级）
+
+在仓库根目录创建 `.aiignore` 文件，用于排除特定文件的索引，这些文件应该被 git-ai 忽略但不一定要被 Git 忽略：
+
+```bash
+# .aiignore 示例
+test-fixtures/**
+*.generated.ts
+docs/api-reference/**
+```
+
+#### `.git-ai/include.txt` - 强制包含（覆盖 `.gitignore`）
+
+有时您需要索引生成的代码或在 `.gitignore` 中但对代码理解很重要的文件。创建 `.git-ai/include.txt` 来强制索引特定模式：
+
+```bash
+# .git-ai/include.txt 示例
+# 包含生成的 API 客户端
+generated/api/**
+
+# 包含特定的构建产物，其中包含重要的类型定义
+dist/types/**
+
+# 包含特定被忽略目录中的代码
+vendor/important-lib/**
+```
+
+**优先级顺序（从高到低）：**
+1. `.aiignore` - 显式排除规则始终生效
+2. `.git-ai/include.txt` - 强制包含模式覆盖 `.gitignore`
+3. `.gitignore` - 标准 Git 忽略模式
+
+**支持的模式语法：**
+- `**` - 匹配任意数量的目录
+- `*` - 匹配目录内的任意字符
+- `directory/` - 匹配整个目录（自动转换为 `directory/**`）
+- `file.ts` - 匹配特定文件
+- 以 `#` 开头的行为注释
+
+**配置示例：**
+
+```bash
+# .gitignore
+dist/
+generated/
+*.log
+
+# .git-ai/include.txt
+generated/api/**
+generated/types/**
+
+# .aiignore (覆盖所有规则)
+generated/test-data/**
+```
+
+此配置下：
+- ✅ `generated/api/client.ts` - 被索引（通过 include.txt 包含）
+- ✅ `generated/types/models.ts` - 被索引（通过 include.txt 包含）
+- ❌ `generated/test-data/mock.ts` - 不被索引（.aiignore 优先级最高）
+- ❌ `dist/bundle.js` - 不被索引（在 .gitignore 中，不在 include.txt 中）
+
+---
+
 ## 🛠️ 故障排除
 
 ### Windows 安装问题

src/core/indexer.ts

@@ -9,6 +9,9 @@ import { ChunkRow, RefRow } from './types';
 import { toPosixPath } from './paths';
 import { getCurrentCommitHash } from './git';
 
+// Supported file extensions for indexing
+const INDEXABLE_EXTENSIONS = 'ts,tsx,js,jsx,java,c,h,go,py,rs,md,mdx,yml,yaml';
+
 export interface IndexOptions {
   repoRoot: string;
   scanRoot?: string;
@@ -36,6 +39,23 @@ async function loadIgnorePatterns(repoRoot: string, fileName: string): Promise<s
     .filter((l): l is string => Boolean(l));
 }
 
+async function loadIncludePatterns(repoRoot: string): Promise<string[]> {
+  const includePath = path.join(repoRoot, '.git-ai', 'include.txt');
+  if (!await fs.pathExists(includePath)) return [];
+  const raw = await fs.readFile(includePath, 'utf-8');
+  return raw
+    .split('\n')
+    .map(l => l.trim())
+    .map((l) => {
+      if (l.length === 0) return null;
+      if (l.startsWith('#')) return null;
+      const withoutLeadingSlash = l.startsWith('/') ? l.slice(1) : l;
+      if (withoutLeadingSlash.endsWith('/')) return `${withoutLeadingSlash}**`;
+      return withoutLeadingSlash;
+    })
+    .filter((l): l is string => Boolean(l));
+}
+
 function inferIndexLang(file: string): IndexLang {
   if (file.endsWith('.md') || file.endsWith('.mdx')) return 'markdown';
   if (file.endsWith('.yml') || file.endsWith('.yaml')) return 'yaml';
@@ -71,30 +91,74 @@ export class IndexerV2 {
 
     const aiIgnore = await loadIgnorePatterns(this.repoRoot, '.aiignore');
     const gitIgnore = await loadIgnorePatterns(this.repoRoot, '.gitignore');
-    const files = await glob('**/*.{ts,tsx,js,jsx,java,c,h,go,py,rs,md,mdx,yml,yaml}', {
+    const includePatterns = await loadIncludePatterns(this.repoRoot);
+
+    // Base ignore patterns that are always applied
+    const baseIgnore = [
+      'node_modules/**',
+      '**/node_modules/**',
+      '.git/**',
+      '**/.git/**',
+      '.git-ai/**',
+      '**/.git-ai/**',
+      '.repo/**',
+      '**/.repo/**',
+      'dist/**',
+      '**/dist/**',
+      'target/**',
+      '**/target/**',
+      'build/**',
+      '**/build/**',
+      '.gradle/**',
+      '**/.gradle/**',
+    ];
+
+    // Get files with normal ignore patterns (aiIgnore and gitIgnore)
+    const filesNormal = await glob(`**/*.{${INDEXABLE_EXTENSIONS}}`, {
       cwd: this.scanRoot,
       nodir: true,
       ignore: [
-        'node_modules/**',
-        '**/node_modules/**',
-        '.git/**',
-        '**/.git/**',
-        '.git-ai/**',
-        '**/.git-ai/**',
-        '.repo/**',
-        '**/.repo/**',
-        'dist/**',
-        'target/**',
-        '**/target/**',
-        'build/**',
-        '**/build/**',
-        '.gradle/**',
-        '**/.gradle/**',
+        ...baseIgnore,
         ...aiIgnore,
         ...gitIgnore,
       ],
     });
 
+    let files = filesNormal;
+
+    // If include patterns exist, also get files matching those patterns (ignoring gitIgnore but respecting aiIgnore)
+    if (includePatterns.length > 0) {
+      // For each include pattern, get files matching it without gitIgnore restrictions
+      const includedFileSets = await Promise.all(
+        includePatterns.map(async (pattern) => {
+          // Ensure pattern covers all file extensions we support
+          let fullPattern = pattern;
+          // If pattern is a directory pattern (e.g., "generated/**"), append file extensions
+          if (pattern.endsWith('**')) {
+            fullPattern = `${pattern}/*.{${INDEXABLE_EXTENSIONS}}`;
+          } else if (!pattern.match(/\.(ts|tsx|js|jsx|java|c|h|go|py|rs|md|mdx|yml|yaml)$/)) {
+            // If pattern doesn't end with a file extension, treat it as a directory
+            fullPattern = `${pattern}/**/*.{${INDEXABLE_EXTENSIONS}}`;
+          }
+
+          return glob(fullPattern, {
+            cwd: this.scanRoot,
+            nodir: true,
+            ignore: [
+              ...baseIgnore,
+              ...aiIgnore,
+              // Note: gitIgnore is NOT applied here
+            ],
+          });
+        })
+      );
+
+      // Flatten and merge with normal files
+      const includedFiles = includedFileSets.flat();
+      const fileSet = new Set([...filesNormal, ...includedFiles]);
+      files = Array.from(fileSet);
+    }
+
     const languages = Array.from(new Set(files.map(inferIndexLang)));
     const { byLang } = await openTablesByLang({
       dbDir,