mcp-vector-search 0.9.3__tar.gz → 0.12.1__tar.gz

{mcp_vector_search-0.9.3 → mcp_vector_search-0.12.1}/.gitignore

@@ -1,3 +1,13 @@
+# All dotfiles and dot-directories (catch-all)
+.*
+!.gitignore
+!.github/
+!.pre-commit-config.yaml
+
+# Explicitly exclude common dotfiles/directories
+.claude/
+.changesets/
+
 # MCP Vector Search specific
 .mcp-vector-search/
 *.db
@@ -179,6 +189,8 @@ test_*.py
 *_test.py
 debug_*.py
 test_js_project/
+test-screenshots/
+chunk-graph.json
 
 # UV specific
 .uv/
@@ -287,3 +299,6 @@ pip-wheel-metadata/
 site/
 temp/
 tmp/
+
+# MCP Ticketer
+.mcp-ticketer/

mcp_vector_search-0.12.1/CLAUDE.md

@@ -0,0 +1,49 @@
+# Project Memory Configuration
+
+This project uses KuzuMemory for intelligent context management.
+
+## Project Information
+- **Path**: /Users/masa/Projects/mcp-vector-search
+- **Language**: Python
+- **Framework**: Unknown
+
+## Memory Integration
+
+KuzuMemory is configured to enhance all AI interactions with project-specific context.
+
+### Available Commands:
+- `kuzu-memory enhance <prompt>` - Enhance prompts with project context
+- `kuzu-memory learn <content>` - Store learning from conversations (async)
+- `kuzu-memory recall <query>` - Query project memories
+- `kuzu-memory stats` - View memory statistics
+
+### MCP Tools Available:
+When interacting with Claude Desktop, the following MCP tools are available:
+- **kuzu_enhance**: Enhance prompts with project memories
+- **kuzu_learn**: Store new learnings asynchronously
+- **kuzu_recall**: Query specific memories
+- **kuzu_stats**: Get memory system statistics
+
+## Project Context
+
+CLI-first semantic code search with MCP integration
+
+## Key Technologies
+- Python
+
+## Development Guidelines
+- Use kuzu-memory enhance for all AI interactions
+- Store important decisions with kuzu-memory learn
+- Query context with kuzu-memory recall when needed
+- Keep memories project-specific and relevant
+
+## Memory Guidelines
+
+- Store project decisions and conventions
+- Record technical specifications and API details
+- Capture user preferences and patterns
+- Document error solutions and workarounds
+
+---
+
+*Generated by KuzuMemory Claude Hooks Installer*

{mcp_vector_search-0.9.3 → mcp_vector_search-0.12.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mcp-vector-search
-Version: 0.9.3
+Version: 0.12.1
 Summary: CLI-first semantic code search with MCP integration
 Project-URL: Homepage, https://github.com/bobmatnyc/mcp-vector-search
 Project-URL: Documentation, https://mcp-vector-search.readthedocs.io
@@ -39,6 +39,7 @@ Classifier: Topic :: Software Development :: Code Generators
 Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Requires-Python: >=3.11
 Requires-Dist: aiofiles>=23.0.0
+Requires-Dist: authlib>=1.6.4
 Requires-Dist: chromadb>=0.5.0
 Requires-Dist: click-didyoumean>=0.3.0
 Requires-Dist: httpx>=0.25.0
@@ -110,49 +111,68 @@ cd mcp-vector-search
 uv sync && uv pip install -e .
 ```
 
-### Complete Setup with Install Command
+### Complete Setup (One Command)
 
-The new **enhanced install command** provides a complete one-step setup:
+The **hierarchical install command** (v0.13.0) provides complete project setup and MCP integration management:
 
 ```bash
-# Interactive setup with MCP configuration
+# Quick setup (recommended)
 mcp-vector-search install
 
-# Setup without MCP configuration
-mcp-vector-search install --no-mcp
+# This will:
+# 1. Initialize your project configuration
+# 2. Automatically index your codebase
+# 3. Provide next-step hints for MCP integration
 
-# Setup for specific MCP tool
-mcp-vector-search install --mcp-tool "Claude Code"
-
-# Setup without automatic indexing
-mcp-vector-search install --no-index
+# Install with all MCP integrations at once
+mcp-vector-search install --with-mcp
 
 # Custom file extensions
 mcp-vector-search install --extensions .py,.js,.ts,.dart
+
+# Skip automatic indexing
+mcp-vector-search install --no-auto-index
 ```
 
-The install command:
-- Initializes your project configuration
-- Detects and configures MCP tools (Claude Code, Cursor, Windsurf, VS Code)
-- Automatically indexes your codebase
-- Provides rich progress indicators and next-step hints
+### Add MCP Integration for AI Tools
 
-### Basic Usage
+```bash
+# Add Claude Code integration (project-scoped)
+mcp-vector-search install claude-code
+
+# Add Cursor IDE integration (global)
+mcp-vector-search install cursor
+
+# Add Claude Desktop integration (global)
+mcp-vector-search install claude-desktop
+
+# See all available platforms
+mcp-vector-search install list
+```
+
+### Remove MCP Integrations
 
 ```bash
-# Initialize your project
-mcp-vector-search init
+# Remove specific platform
+mcp-vector-search uninstall claude-code
 
-# Index your codebase
-mcp-vector-search index
+# Remove all integrations
+mcp-vector-search uninstall --all
 
+# List configured integrations
+mcp-vector-search uninstall list
+```
+
+### Basic Usage
+
+```bash
 # Search your code
 mcp-vector-search search "authentication logic"
 mcp-vector-search search "database connection setup"
 mcp-vector-search search "error handling patterns"
 
-# Setup automatic reindexing (recommended)
-mcp-vector-search auto-index setup --method all
+# Index your codebase (if not done during install)
+mcp-vector-search index
 
 # Check project status
 mcp-vector-search status
@@ -191,7 +211,50 @@ See [docs/VERSIONING_WORKFLOW.md](docs/VERSIONING_WORKFLOW.md) for complete docu
 
 ### Commands
 
-#### `init` - Initialize Project
+#### `install` - Install Project and MCP Integrations (v0.13.0)
+```bash
+# Quick setup (recommended)
+mcp-vector-search install
+
+# Install with all MCP integrations
+mcp-vector-search install --with-mcp
+
+# Custom file extensions
+mcp-vector-search install --extensions .py,.js,.ts
+
+# Skip automatic indexing
+mcp-vector-search install --no-auto-index
+
+# Platform-specific MCP integration
+mcp-vector-search install claude-code      # Project-scoped
+mcp-vector-search install claude-desktop   # Global
+mcp-vector-search install cursor           # Global
+mcp-vector-search install windsurf         # Global
+mcp-vector-search install vscode           # Global
+
+# List available platforms
+mcp-vector-search install list
+```
+
+#### `uninstall` - Remove MCP Integrations (v0.13.0)
+```bash
+# Remove specific platform
+mcp-vector-search uninstall claude-code
+
+# Remove all integrations
+mcp-vector-search uninstall --all
+
+# List configured integrations
+mcp-vector-search uninstall list
+
+# Skip backup creation
+mcp-vector-search uninstall claude-code --no-backup
+
+# Alias (same as uninstall)
+mcp-vector-search remove claude-code
+```
+
+#### `init` - Initialize Project (Simple)
 ```bash
 # Basic initialization
 mcp-vector-search init

{mcp_vector_search-0.9.3 → mcp_vector_search-0.12.1}/README.md

@@ -53,49 +53,68 @@ cd mcp-vector-search
 uv sync && uv pip install -e .
 ```
 
-### Complete Setup with Install Command
+### Complete Setup (One Command)
 
-The new **enhanced install command** provides a complete one-step setup:
+The **hierarchical install command** (v0.13.0) provides complete project setup and MCP integration management:
 
 ```bash
-# Interactive setup with MCP configuration
+# Quick setup (recommended)
 mcp-vector-search install
 
-# Setup without MCP configuration
-mcp-vector-search install --no-mcp
+# This will:
+# 1. Initialize your project configuration
+# 2. Automatically index your codebase
+# 3. Provide next-step hints for MCP integration
 
-# Setup for specific MCP tool
-mcp-vector-search install --mcp-tool "Claude Code"
-
-# Setup without automatic indexing
-mcp-vector-search install --no-index
+# Install with all MCP integrations at once
+mcp-vector-search install --with-mcp
 
 # Custom file extensions
 mcp-vector-search install --extensions .py,.js,.ts,.dart
+
+# Skip automatic indexing
+mcp-vector-search install --no-auto-index
 ```
 
-The install command:
-- Initializes your project configuration
-- Detects and configures MCP tools (Claude Code, Cursor, Windsurf, VS Code)
-- Automatically indexes your codebase
-- Provides rich progress indicators and next-step hints
+### Add MCP Integration for AI Tools
 
-### Basic Usage
+```bash
+# Add Claude Code integration (project-scoped)
+mcp-vector-search install claude-code
+
+# Add Cursor IDE integration (global)
+mcp-vector-search install cursor
+
+# Add Claude Desktop integration (global)
+mcp-vector-search install claude-desktop
+
+# See all available platforms
+mcp-vector-search install list
+```
+
+### Remove MCP Integrations
 
 ```bash
-# Initialize your project
-mcp-vector-search init
+# Remove specific platform
+mcp-vector-search uninstall claude-code
 
-# Index your codebase
-mcp-vector-search index
+# Remove all integrations
+mcp-vector-search uninstall --all
 
+# List configured integrations
+mcp-vector-search uninstall list
+```
+
+### Basic Usage
+
+```bash
 # Search your code
 mcp-vector-search search "authentication logic"
 mcp-vector-search search "database connection setup"
 mcp-vector-search search "error handling patterns"
 
-# Setup automatic reindexing (recommended)
-mcp-vector-search auto-index setup --method all
+# Index your codebase (if not done during install)
+mcp-vector-search index
 
 # Check project status
 mcp-vector-search status
@@ -134,7 +153,50 @@ See [docs/VERSIONING_WORKFLOW.md](docs/VERSIONING_WORKFLOW.md) for complete docu
 
 ### Commands
 
-#### `init` - Initialize Project
+#### `install` - Install Project and MCP Integrations (v0.13.0)
+```bash
+# Quick setup (recommended)
+mcp-vector-search install
+
+# Install with all MCP integrations
+mcp-vector-search install --with-mcp
+
+# Custom file extensions
+mcp-vector-search install --extensions .py,.js,.ts
+
+# Skip automatic indexing
+mcp-vector-search install --no-auto-index
+
+# Platform-specific MCP integration
+mcp-vector-search install claude-code      # Project-scoped
+mcp-vector-search install claude-desktop   # Global
+mcp-vector-search install cursor           # Global
+mcp-vector-search install windsurf         # Global
+mcp-vector-search install vscode           # Global
+
+# List available platforms
+mcp-vector-search install list
+```
+
+#### `uninstall` - Remove MCP Integrations (v0.13.0)
+```bash
+# Remove specific platform
+mcp-vector-search uninstall claude-code
+
+# Remove all integrations
+mcp-vector-search uninstall --all
+
+# List configured integrations
+mcp-vector-search uninstall list
+
+# Skip backup creation
+mcp-vector-search uninstall claude-code --no-backup
+
+# Alias (same as uninstall)
+mcp-vector-search remove claude-code
+```
+
+#### `init` - Initialize Project (Simple)
 ```bash
 # Basic initialization
 mcp-vector-search init

{mcp_vector_search-0.9.3 → mcp_vector_search-0.12.1}/docs/CHANGELOG.md

@@ -7,6 +7,61 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.13.0] - 2025-10-27
+
+### Added
+- **Hierarchical Install/Uninstall Commands**: Complete CLI command restructuring for better user experience
+  - New `install` command with platform-specific subcommands
+  - New `uninstall` command for removing MCP integrations
+  - Support for 5 MCP platforms: claude-code, claude-desktop, cursor, windsurf, vscode
+  - Project-scoped configuration (`.mcp.json`) for claude-code
+  - Global-scoped configuration for other platforms (home directory)
+  - `install list` and `uninstall list` for platform discovery
+  - `remove` command as alias for `uninstall`
+  - `--with-mcp` flag to install all MCP integrations at once
+  - `--no-auto-index` flag to skip automatic indexing
+  - `--no-backup` flag for uninstall operations
+
+### Changed
+- **MCP Command Reserved**: The `mcp` command is now exclusively for MCP server operations (not installation)
+  - `mcp-vector-search mcp` - Start MCP server (stdio mode)
+  - `mcp-vector-search mcp list` - List configured servers
+  - `mcp-vector-search mcp test` - Test server startup
+- **Install Command Reintroduced**: Previously deprecated `install` command is back with new hierarchical structure
+  - `install` is now the primary command for project setup and MCP integration
+  - `init` remains available for simple project initialization without MCP
+
+### Deprecated
+- `init-mcp` command deprecated in favor of `install claude-code` (or other platform subcommands)
+
+### Files Modified
+- `src/mcp_vector_search/cli/commands/install.py` - Rewritten with hierarchical structure
+- `src/mcp_vector_search/cli/commands/uninstall.py` - New command for MCP removal
+- `src/mcp_vector_search/cli/main.py` - Updated command registration and help text
+- `CLAUDE.md` - Updated with new command examples and structure
+- `README.md` - Updated quick start and command documentation
+
+### Migration Guide
+**Old Pattern** (v0.12.0 and earlier):
+```bash
+mcp-vector-search init --mcp
+```
+
+**New Pattern** (v0.13.0+):
+```bash
+# Quick setup
+mcp-vector-search install
+
+# Add specific platform
+mcp-vector-search install claude-code
+```
+
+The new structure provides:
+- Better command discoverability
+- Fine-grained platform control
+- Clear separation between installation and MCP server operations
+- Ability to remove specific integrations
+
 ### Added
 - **HTML Language Support**: Full HTML parser implementation
   - Semantic content extraction from HTML tags
@@ -56,6 +111,52 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Created parser utilities module to reduce code duplication (potential -800 to -1000 LOC)
 - Proper LRU cache implementation with statistics and configurable size
 
+## [0.12.0] - 2025-10-25
+
+### Added
+- **Hierarchical Graph Visualization**: Complete directory, file, and chunk-based visualization system
+  - Interactive D3.js force-directed graph with expand/collapse functionality
+  - Directory nodes expand to show all files and subdirectories
+  - File nodes expand to show individual code chunks with metadata
+  - Visual hierarchy with optimized 35px folder/file icons
+  - Cache-busting URL parameters and meta tags for fresh data loading
+
+### Enhanced
+- **Graph Link Visibility**: Improved visual clarity of relationships
+  - Enhanced link colors with better opacity (0.4 for directories, 0.3 for files)
+  - Dynamic link distances based on node types (150px directories, 80px files/chunks)
+  - Optimized collision detection with dynamic radius calculations
+  - Better auto-spacing with enhanced force simulation parameters
+
+- **Expand/Collapse Functionality**: Robust node interaction system
+  - Working directory expansion showing all child files and subdirectories
+  - File expansion revealing all contained code chunks
+  - Preserves original link structure before D3 modifications
+  - Proper state management for expanded/collapsed nodes
+  - Visual feedback with node color changes
+
+### Fixed
+- **Parent Directory Linking**: Corrected absolute to relative path conversion
+  - Fixed lookup of parent directory IDs during graph generation
+  - Proper handling of nested directory structures
+  - Consistent path normalization across all node types
+
+- **Stale Data Prevention**: Enhanced data freshness mechanisms
+  - Added `Cache-Control: no-cache, no-store, must-revalidate` meta tags
+  - Implemented timestamp-based cache-busting for JSON files
+  - Ensured visualization directory contains latest graph data
+  - Prevents browser caching of outdated graph structures
+
+### Files Modified
+- `src/mcp_vector_search/visualization/index.html` - UI improvements, expand/collapse fixes, cache prevention
+- `src/mcp_vector_search/cli/commands/visualize.py` - Parent directory path normalization and linking fixes
+
+### Technical Details
+- Zero new dependencies (uses existing D3.js v7)
+- Enhanced force simulation with configurable parameters
+- Improved node interaction state management
+- Better visual design with optimized spacing and colors
+
 ## [0.5.0] - 2025-10-02
 
 ### Added

{mcp_vector_search-0.9.3 → mcp_vector_search-0.12.1}/pyproject.toml

@@ -38,6 +38,7 @@ dependencies = [
     "mcp>=1.12.4",
     "click-didyoumean>=0.3.0",
     "packaging>=23.0",
+    "authlib>=1.6.4",
 ]
 
 [dependency-groups]
@@ -47,6 +48,8 @@ dev = [
     "pytest-benchmark>=4.0.0",
     "pytest-cov>=4.1.0",
     "pytest-mock>=3.11.0",
+    "pytest-xdist>=3.3.0",
+    "pytest-watch>=4.2.0",
     "hypothesis>=6.88.0",
     "black>=23.0.0",
     "ruff>=0.1.0",
@@ -80,13 +83,7 @@ path = "src/mcp_vector_search/__init__.py"
 packages = ["src/mcp_vector_search"]
 
 # UV-specific configuration
-[tool.uv]
-dev-dependencies = [
-    "pytest-xdist>=3.3.0",
-    "pytest-watch>=4.2.0",
-]
-
-# UV sources configuration removed - using standard dependencies
+# Note: dev-dependencies moved to [dependency-groups] section above
 
 # Development tools configuration
 [tool.black]

{mcp_vector_search-0.9.3 → mcp_vector_search-0.12.1}/pytest.ini

@@ -33,7 +33,8 @@ minversion = 6.0
 timeout = 300
 
 # Async test configuration
-asyncio_mode = auto
+asyncio_mode = "auto"
+asyncio_default_fixture_loop_scope = function
 
 # Logging configuration
 log_cli = false

{mcp_vector_search-0.9.3 → mcp_vector_search-0.12.1}/src/mcp_vector_search/__init__.py

@@ -1,7 +1,7 @@
 """MCP Vector Search - CLI-first semantic code search with MCP integration."""
 
-__version__ = "0.9.3"
-__build__ = "40"
+__version__ = "0.12.1"
+__build__ = "51"
 __author__ = "Robert Matsuoka"
 __email__ = "bobmatnyc@gmail.com"
 

{mcp_vector_search-0.9.3 → mcp_vector_search-0.12.1}/src/mcp_vector_search/cli/commands/index.py

@@ -67,6 +67,13 @@ def main(
         max=128,
         rich_help_panel="⚡ Performance",
     ),
+    debug: bool = typer.Option(
+        False,
+        "--debug",
+        "-d",
+        help="Enable debug output (shows hierarchy building details)",
+        rich_help_panel="🔍 Debugging",
+    ),
 ) -> None:
     """📑 Index your codebase for semantic search.
 
@@ -114,6 +121,7 @@ def main(
                 force_reindex=force,
                 batch_size=batch_size,
                 show_progress=True,
+                debug=debug,
             )
         )
 
@@ -134,6 +142,7 @@ async def run_indexing(
     force_reindex: bool = False,
     batch_size: int = 32,
     show_progress: bool = True,
+    debug: bool = False,
 ) -> None:
     """Run the indexing process."""
     # Load project configuration
@@ -179,6 +188,7 @@ async def run_indexing(
         database=database,
         project_root=project_root,
         file_extensions=file_extensions,
+        debug=debug,
     )
 
     try:
@@ -324,19 +334,41 @@ async def _run_batch_indexing(
                         )
                     )
 
+            # Rebuild directory index after indexing completes
+            try:
+                import os
+
+                chunk_stats = {}
+                for file_path in files_to_index:
+                    try:
+                        mtime = os.path.getmtime(file_path)
+                        chunk_stats[str(file_path)] = {
+                            "modified": mtime,
+                            "chunks": 1,  # Placeholder - real counts are in database
+                        }
+                    except OSError:
+                        pass
+
+                indexer.directory_index.rebuild_from_files(
+                    files_to_index, indexer.project_root, chunk_stats=chunk_stats
+                )
+                indexer.directory_index.save()
+            except Exception as e:
+                logger.error(f"Failed to update directory index: {e}")
+
             # Final progress summary
             console.print()
             if failed_count > 0:
                 console.print(
                     f"[yellow]⚠ {failed_count} files failed to index[/yellow]"
                 )
-                error_log_path = indexer.project_root / ".mcp-vector-search" / "indexing_errors.log"
+                error_log_path = (
+                    indexer.project_root / ".mcp-vector-search" / "indexing_errors.log"
+                )
                 if error_log_path.exists():
                     # Prune log to keep only last 1000 errors
                     _prune_error_log(error_log_path, max_lines=1000)
-                    console.print(
-                        f"[dim]  → See details in: {error_log_path}[/dim]"
-                    )
+                    console.print(f"[dim]  → See details in: {error_log_path}[/dim]")
     else:
         # Non-progress mode (fallback to original behavior)
         indexed_count = await indexer.index_project(
@@ -643,22 +675,10 @@ def watch_cmd(
     watch_app()
 
 
-@index_app.command("auto")
-def auto_cmd() -> None:
-    """🔄 Manage automatic indexing.
-
-    Configure automatic indexing strategies like git hooks and scheduled tasks.
-    This command provides subcommands for setup, status, and checking.
-
-    Examples:
-        mcp-vector-search index auto setup
-        mcp-vector-search index auto status
-        mcp-vector-search index auto check
-    """
-    from .auto_index import auto_index_app
+# Import and register auto-index sub-app as a proper typer group
+from .auto_index import auto_index_app  # noqa: E402
 
-    # This will show help for the auto subcommands
-    auto_index_app()
+index_app.add_typer(auto_index_app, name="auto", help="🔄 Manage automatic indexing")
 
 
 @index_app.command("health")
@@ -702,17 +722,19 @@ def _prune_error_log(log_path: Path, max_lines: int = 1000) -> None:
         max_lines: Maximum number of lines to keep (default: 1000)
     """
     try:
-        with open(log_path, 'r') as f:
+        with open(log_path) as f:
             lines = f.readlines()
 
         if len(lines) > max_lines:
             # Keep only the last max_lines lines
             pruned_lines = lines[-max_lines:]
 
-            with open(log_path, 'w') as f:
+            with open(log_path, "w") as f:
                 f.writelines(pruned_lines)
 
-            logger.debug(f"Pruned error log from {len(lines)} to {len(pruned_lines)} lines")
+            logger.debug(
+                f"Pruned error log from {len(lines)} to {len(pruned_lines)} lines"
+            )
     except Exception as e:
         logger.warning(f"Failed to prune error log: {e}")