fix: keep backward-compat

Previous fix revealed the real issue: embedding server fails to start within 120s,
not timeout issues. The error was hidden because both stdout and stderr were
redirected to DEVNULL in CI.

Changes:
- Keep stderr output in CI environment for debugging
- Only redirect stdout to DEVNULL to avoid buffer deadlock
- This will help us see why embedding server startup is failing

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

.gitattributes

@@ -0,0 +1 @@
+paper_plot/data/big_graph_degree_data.npz filter=lfs diff=lfs merge=lfs -text

.github/workflows/build-reusable.yml

@@ -87,7 +87,7 @@ jobs:
     runs-on: ${{ matrix.os }}
 
     steps:
-      - uses: actions/checkout@v5
+      - uses: actions/checkout@v4
         with:
           ref: ${{ inputs.ref }}
           submodules: recursive
@@ -98,23 +98,21 @@ jobs:
           python-version: ${{ matrix.python }}
 
       - name: Install uv
-        uses: astral-sh/setup-uv@v6
+        uses: astral-sh/setup-uv@v4
 
       - name: Install system dependencies (Ubuntu)
         if: runner.os == 'Linux'
         run: |
           sudo apt-get update
           sudo apt-get install -y libomp-dev libboost-all-dev protobuf-compiler libzmq3-dev \
-            pkg-config libabsl-dev libaio-dev libprotobuf-dev \
-            patchelf
+            pkg-config libopenblas-dev patchelf libabsl-dev libaio-dev libprotobuf-dev
 
           # Install Intel MKL for DiskANN
           wget -q https://registrationcenter-download.intel.com/akdlm/IRC_NAS/79153e0f-74d7-45af-b8c2-258941adf58a/intel-onemkl-2025.0.0.940.sh
           sudo sh intel-onemkl-2025.0.0.940.sh -a --components intel.oneapi.lin.mkl.devel --action install --eula accept -s
           source /opt/intel/oneapi/setvars.sh
           echo "MKLROOT=/opt/intel/oneapi/mkl/latest" >> $GITHUB_ENV
-          echo "LD_LIBRARY_PATH=/opt/intel/oneapi/compiler/latest/linux/compiler/lib/intel64_lin" >> $GITHUB_ENV
-          echo "LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/opt/intel/oneapi/mkl/latest/lib/intel64" >> $GITHUB_ENV
+          echo "LD_LIBRARY_PATH=/opt/intel/oneapi/mkl/latest/lib/intel64:$LD_LIBRARY_PATH" >> $GITHUB_ENV
 
       - name: Install system dependencies (macOS)
         if: runner.os == 'macOS'
@@ -306,53 +304,3 @@ jobs:
         with:
           name: packages-${{ matrix.os }}-py${{ matrix.python }}
           path: packages/*/dist/
-
-
-  arch-smoke:
-    name: Arch Linux smoke test (install & import)
-    needs: build
-    runs-on: ubuntu-latest
-    container:
-      image: archlinux:latest
-
-    steps:
-      - name: Prepare system
-        run: |
-          pacman -Syu --noconfirm
-          pacman -S --noconfirm python python-pip gcc git zlib openssl
-
-      - name: Download ALL wheel artifacts from this run
-        uses: actions/download-artifact@v5
-        with:
-          # Don't specify name, download all artifacts
-          path: ./wheels
-
-      - name: Install uv
-        uses: astral-sh/setup-uv@v6
-
-      - name: Create virtual environment and install wheels
-        run: |
-          uv venv
-          source .venv/bin/activate || source .venv/Scripts/activate
-          uv pip install --find-links wheels leann-core
-          uv pip install --find-links wheels leann-backend-hnsw
-          uv pip install --find-links wheels leann-backend-diskann
-          uv pip install --find-links wheels leann
-
-      - name: Import & tiny runtime check
-        env:
-          OMP_NUM_THREADS: 1
-          MKL_NUM_THREADS: 1
-        run: |
-          source .venv/bin/activate || source .venv/Scripts/activate
-          python - <<'PY'
-          import leann
-          import leann_backend_hnsw as h
-          import leann_backend_diskann as d
-          from leann import LeannBuilder, LeannSearcher
-          b = LeannBuilder(backend_name="hnsw")
-          b.add_text("hello arch")
-          b.build_index("arch_demo.leann")
-          s = LeannSearcher("arch_demo.leann")
-          print("search:", s.search("hello", top_k=1))
-          PY

.github/workflows/link-check.yml

@@ -14,6 +14,6 @@ jobs:
       - uses: actions/checkout@v4
       - uses: lycheeverse/lychee-action@v2
         with:
-          args: --no-progress --insecure --user-agent 'curl/7.68.0' README.md docs/ apps/ examples/ benchmarks/
+          args: --no-progress --insecure README.md docs/ apps/ examples/ benchmarks/
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.gitignore

@@ -18,7 +18,6 @@ demo/experiment_results/**/*.json
 *.eml
 *.emlx
 *.json
-!.vscode/*.json
 *.sh
 *.txt
 !CMakeLists.txt

.vscode/extensions.json

@@ -1,5 +0,0 @@
-{
-    "recommendations": [
-        "charliermarsh.ruff",
-    ]
-}

.vscode/settings.json

@@ -1,22 +0,0 @@
-{
-  "python.defaultInterpreterPath": ".venv/bin/python",
-  "python.terminal.activateEnvironment": true,
-  "[python]": {
-    "editor.defaultFormatter": "charliermarsh.ruff",
-    "editor.formatOnSave": true,
-    "editor.codeActionsOnSave": {
-      "source.organizeImports": "explicit",
-      "source.fixAll": "explicit"
-    },
-    "editor.insertSpaces": true,
-    "editor.tabSize": 4
-  },
-  "ruff.enable": true,
-  "files.watcherExclude": {
-    "**/.venv/**": true,
-    "**/__pycache__/**": true,
-    "**/*.egg-info/**": true,
-    "**/build/**": true,
-    "**/dist/**": true
-  }
-}

README.md

@@ -5,7 +5,7 @@
 <p align="center">
   <img src="https://img.shields.io/badge/Python-3.9%20%7C%203.10%20%7C%203.11%20%7C%203.12%20%7C%203.13-blue.svg" alt="Python Versions">
   <img src="https://github.com/yichuan-w/LEANN/actions/workflows/build-and-publish.yml/badge.svg" alt="CI Status">
-  <img src="https://img.shields.io/badge/Platform-Ubuntu%20%26%20Arch%20%26%20WSL%20%7C%20macOS%20(ARM64%2FIntel)-lightgrey" alt="Platform">
+  <img src="https://img.shields.io/badge/Platform-Ubuntu%20%7C%20macOS%20(ARM64%2FIntel)-lightgrey" alt="Platform">
   <img src="https://img.shields.io/badge/License-MIT-green.svg" alt="MIT License">
   <img src="https://img.shields.io/badge/MCP-Native%20Integration-blue" alt="MCP Integration">
 </p>
@@ -31,7 +31,7 @@ LEANN achieves this through *graph-based selective recomputation* with *high-deg
   <img src="assets/effects.png" alt="LEANN vs Traditional Vector DB Storage Comparison" width="70%">
 </p>
 
-> **The numbers speak for themselves:** Index 60 million text chunks in just 6GB instead of 201GB. From emails to browser history, everything fits on your laptop. [See detailed benchmarks for different applications below ↓](#-storage-comparison)
+> **The numbers speak for themselves:** Index 60 million text chunks in just 6GB instead of 201GB. From emails to browser history, everything fits on your laptop. [See detailed benchmarks for different applications below ↓](#storage-comparison)
 
 
 🔒 **Privacy:** Your data never leaves your laptop. No OpenAI, no cloud, no "terms of service".
@@ -70,8 +70,6 @@ uv venv
 source .venv/bin/activate
 uv pip install leann
 ```
-<!--
-> Low-resource? See “Low-resource setups” in the [Configuration Guide](docs/configuration-guide.md#low-resource-setups). -->
 
 <details>
 <summary>
@@ -87,60 +85,15 @@ git submodule update --init --recursive
 ```
 
 **macOS:**
-
-Note: DiskANN requires MacOS 13.3 or later.
-
 ```bash
-brew install libomp boost protobuf zeromq pkgconf
-uv sync --extra diskann
+brew install llvm libomp boost protobuf zeromq pkgconf
+CC=$(brew --prefix llvm)/bin/clang CXX=$(brew --prefix llvm)/bin/clang++ uv sync
 ```
 
-**Linux (Ubuntu/Debian):**
-
-Note: On Ubuntu 20.04, you may need to build a newer Abseil and pin Protobuf (e.g., v3.20.x) for building DiskANN. See [Issue #30](https://github.com/yichuan-w/LEANN/issues/30) for a step-by-step note.
-
-You can manually install [Intel oneAPI MKL](https://www.intel.com/content/www/us/en/developer/tools/oneapi/onemkl.html) instead of `libmkl-full-dev` for DiskANN. You can also use `libopenblas-dev` for building HNSW only, by removing `--extra diskann` in the command below.
-
+**Linux:**
 ```bash
-sudo apt-get update && sudo apt-get install -y \
-  libomp-dev libboost-all-dev protobuf-compiler libzmq3-dev \
-  pkg-config libabsl-dev libaio-dev libprotobuf-dev \
-  libmkl-full-dev
-
-uv sync --extra diskann
-```
-
-**Linux (Arch Linux):**
-
-```bash
-sudo pacman -Syu && sudo pacman -S --needed base-devel cmake pkgconf git gcc \
-  boost boost-libs protobuf abseil-cpp libaio zeromq
-
-# For MKL in DiskANN
-sudo pacman -S --needed base-devel git
-git clone https://aur.archlinux.org/paru-bin.git
-cd paru-bin && makepkg -si
-paru -S intel-oneapi-mkl intel-oneapi-compiler
-source /opt/intel/oneapi/setvars.sh
-
-uv sync --extra diskann
-```
-
-**Linux (RHEL / CentOS Stream / Oracle / Rocky / AlmaLinux):**
-
-See [Issue #50](https://github.com/yichuan-w/LEANN/issues/50) for more details.
-
-```bash
-sudo dnf groupinstall -y "Development Tools"
-sudo dnf install -y libomp-devel boost-devel protobuf-compiler protobuf-devel \
-  abseil-cpp-devel libaio-devel zeromq-devel pkgconf-pkg-config
-
-# For MKL in DiskANN
-sudo dnf install -y intel-oneapi-mkl intel-oneapi-mkl-devel \
-  intel-oneapi-openmp || sudo dnf install -y intel-oneapi-compiler
-source /opt/intel/oneapi/setvars.sh
-
-uv sync --extra diskann
+sudo apt-get install libomp-dev libboost-all-dev protobuf-compiler libabsl-dev libmkl-full-dev libaio-dev libzmq3-dev
+uv sync
 ```
 
 </details>
@@ -231,34 +184,34 @@ All RAG examples share these common parameters. **Interactive mode** is availabl
 
 ```bash
 # Core Parameters (General preprocessing for all examples)
---index-dir DIR              # Directory to store the index (default: current directory)
---query "YOUR QUESTION"      # Single query mode. Omit for interactive chat (type 'quit' to exit), and now you can play with your index interactively
---max-items N                # Limit data preprocessing (default: -1, process all data)
---force-rebuild              # Force rebuild index even if it exists
+--index-dir DIR          # Directory to store the index (default: current directory)
+--query "YOUR QUESTION"  # Single query mode. Omit for interactive chat (type 'quit' to exit), and now you can play with your index interactively
+--max-items N           # Limit data preprocessing (default: -1, process all data)
+--force-rebuild         # Force rebuild index even if it exists
 
 # Embedding Parameters
---embedding-model MODEL      # e.g., facebook/contriever, text-embedding-3-small, mlx-community/Qwen3-Embedding-0.6B-8bit or nomic-embed-text
---embedding-mode MODE        # sentence-transformers, openai, mlx, or ollama
+--embedding-model MODEL  # e.g., facebook/contriever, text-embedding-3-small, nomic-embed-text,mlx-community/Qwen3-Embedding-0.6B-8bit or nomic-embed-text
+--embedding-mode MODE    # sentence-transformers, openai, mlx, or ollama
 
 # LLM Parameters (Text generation models)
---llm TYPE                   # LLM backend: openai, ollama, or hf (default: openai)
---llm-model MODEL            # Model name (default: gpt-4o) e.g., gpt-4o-mini, llama3.2:1b, Qwen/Qwen2.5-1.5B-Instruct
---thinking-budget LEVEL      # Thinking budget for reasoning models: low/medium/high (supported by o3, o3-mini, GPT-Oss:20b, and other reasoning models)
+--llm TYPE              # LLM backend: openai, ollama, or hf (default: openai)
+--llm-model MODEL       # Model name (default: gpt-4o) e.g., gpt-4o-mini, llama3.2:1b, Qwen/Qwen2.5-1.5B-Instruct
+--thinking-budget LEVEL # Thinking budget for reasoning models: low/medium/high (supported by o3, o3-mini, GPT-Oss:20b, and other reasoning models)
 
 # Search Parameters
---top-k N                    # Number of results to retrieve (default: 20)
---search-complexity N        # Search complexity for graph traversal (default: 32)
+--top-k N               # Number of results to retrieve (default: 20)
+--search-complexity N   # Search complexity for graph traversal (default: 32)
 
 # Chunking Parameters
---chunk-size N               # Size of text chunks (default varies by source: 256 for most, 192 for WeChat)
---chunk-overlap N            # Overlap between chunks (default varies: 25-128 depending on source)
+--chunk-size N          # Size of text chunks (default varies by source: 256 for most, 192 for WeChat)
+--chunk-overlap N       # Overlap between chunks (default varies: 25-128 depending on source)
 
 # Index Building Parameters
---backend-name NAME          # Backend to use: hnsw or diskann (default: hnsw)
---graph-degree N             # Graph degree for index construction (default: 32)
---build-complexity N         # Build complexity for index construction (default: 64)
---compact / --no-compact     # Use compact storage (default: true). Must be `no-compact` for `no-recompute` build.
---recompute / --no-recompute # Enable/disable embedding recomputation (default: enabled). Should not do a `no-recompute` search in a `recompute` build.
+--backend-name NAME     # Backend to use: hnsw or diskann (default: hnsw)
+--graph-degree N        # Graph degree for index construction (default: 32)
+--build-complexity N    # Build complexity for index construction (default: 64)
+--no-compact           # Disable compact index storage (compact storage IS enabled to save storage by default)
+--no-recompute         # Disable embedding recomputation (recomputation IS enabled to save storage by default)
 ```
 
 </details>
@@ -471,21 +424,21 @@ Once the index is built, you can ask questions like:
 **The future of code assistance is here.** Transform your development workflow with LEANN's native MCP integration for Claude Code. Index your entire codebase and get intelligent code assistance directly in your IDE.
 
 **Key features:**
-- 🔍 **Semantic code search** across your entire project, fully local index and lightweight
+- 🔍 **Semantic code search** across your entire project
 - 📚 **Context-aware assistance** for debugging and development
 - 🚀 **Zero-config setup** with automatic language detection
 
 ```bash
 # Install LEANN globally for MCP integration
-uv tool install leann-core --with leann
-claude mcp add --scope user leann-server -- leann_mcp
+uv tool install leann-core
+
 # Setup is automatic - just start using Claude Code!
 ```
 Try our fully agentic pipeline with auto query rewriting, semantic search planning, and more:
 
 ![LEANN MCP Integration](assets/mcp_leann.png)
 
-**🔥 Ready to supercharge your coding?** [Complete Setup Guide →](packages/leann-mcp/README.md)
+**Ready to supercharge your coding?** [Complete Setup Guide →](packages/leann-mcp/README.md)
 
 ## 🖥️ Command Line Interface
 
@@ -502,8 +455,7 @@ leann --help
 **To make it globally available:**
 ```bash
 # Install the LEANN CLI globally using uv tool
-uv tool install leann-core --with leann
-
+uv tool install leann-core
 
 # Now you can use leann from anywhere without activating venv
 leann --help
@@ -527,35 +479,30 @@ leann ask my-docs --interactive
 
 # List all your indexes
 leann list
-
-# Remove an index
-leann remove my-docs
 ```
 
 **Key CLI features:**
-- Auto-detects document formats (PDF, TXT, MD, DOCX, PPTX + code files)
+- Auto-detects document formats (PDF, TXT, MD, DOCX)
 - Smart text chunking with overlap
 - Multiple LLM providers (Ollama, OpenAI, HuggingFace)
-- Organized index storage in `.leann/indexes/` (project-local)
+- Organized index storage in `~/.leann/indexes/`
 - Support for advanced search parameters
 
 <details>
 <summary><strong>📋 Click to expand: Complete CLI Reference</strong></summary>
 
-You can use `leann --help`, or `leann build --help`, `leann search --help`, `leann ask --help`, `leann list --help`, `leann remove --help` to get the complete CLI reference.
-
 **Build Command:**
 ```bash
-leann build INDEX_NAME --docs DIRECTORY|FILE [DIRECTORY|FILE ...] [OPTIONS]
+leann build INDEX_NAME --docs DIRECTORY [OPTIONS]
 
 Options:
   --backend {hnsw,diskann}     Backend to use (default: hnsw)
   --embedding-model MODEL      Embedding model (default: facebook/contriever)
-  --graph-degree N             Graph degree (default: 32)
-  --complexity N               Build complexity (default: 64)
-  --force                      Force rebuild existing index
-  --compact / --no-compact     Use compact storage (default: true). Must be `no-compact` for `no-recompute` build.
-  --recompute / --no-recompute Enable recomputation (default: true)
+  --graph-degree N            Graph degree (default: 32)
+  --complexity N              Build complexity (default: 64)
+  --force                     Force rebuild existing index
+  --compact                   Use compact storage (default: true)
+  --recompute                 Enable recomputation (default: true)
 ```
 
 **Search Command:**
@@ -563,9 +510,9 @@ Options:
 leann search INDEX_NAME QUERY [OPTIONS]
 
 Options:
-  --top-k N                     Number of results (default: 5)
-  --complexity N                Search complexity (default: 64)
-  --recompute / --no-recompute  Enable/disable embedding recomputation (default: enabled). Should not do a `no-recompute` search in a `recompute` build.
+  --top-k N                   Number of results (default: 5)
+  --complexity N              Search complexity (default: 64)
+  --recompute-embeddings      Use recomputation for highest accuracy
   --pruning-strategy {global,local,proportional}
 ```
 
@@ -580,31 +527,6 @@ Options:
   --top-k N                  Retrieval count (default: 20)
 ```
 
-**List Command:**
-```bash
-leann list
-
-# Lists all indexes across all projects with status indicators:
-# ✅ - Index is complete and ready to use
-# ❌ - Index is incomplete or corrupted
-# 📁 - CLI-created index (in .leann/indexes/)
-# 📄 - App-created index (*.leann.meta.json files)
-```
-
-**Remove Command:**
-```bash
-leann remove INDEX_NAME [OPTIONS]
-
-Options:
-  --force, -f    Force removal without confirmation
-
-# Smart removal: automatically finds and safely removes indexes
-# - Shows all matching indexes across projects
-# - Requires confirmation for cross-project removal
-# - Interactive selection when multiple matches found
-# - Supports both CLI and app-created indexes
-```
-
 </details>
 
 ## 🏗️ Architecture & How It Works

apps/base_rag_example.py

@@ -10,7 +10,6 @@ from typing import Any
 
 import dotenv
 from leann.api import LeannBuilder, LeannChat
-from leann.registry import register_project_directory
 from llama_index.core.node_parser import SentenceSplitter
 
 dotenv.load_dotenv()
@@ -70,14 +69,14 @@ class BaseRAGExample(ABC):
             "--embedding-model",
             type=str,
             default=embedding_model_default,
-            help=f"Embedding model to use (default: {embedding_model_default}), we provide facebook/contriever, text-embedding-3-small,mlx-community/Qwen3-Embedding-0.6B-8bit or nomic-embed-text",
+            help=f"Embedding model to use (default: {embedding_model_default})",
         )
         embedding_group.add_argument(
             "--embedding-mode",
             type=str,
             default="sentence-transformers",
             choices=["sentence-transformers", "openai", "mlx", "ollama"],
-            help="Embedding backend mode (default: sentence-transformers), we provide sentence-transformers, openai, mlx, or ollama",
+            help="Embedding backend mode (default: sentence-transformers)",
         )
 
         # LLM parameters
@@ -87,13 +86,13 @@ class BaseRAGExample(ABC):
             type=str,
             default="openai",
             choices=["openai", "ollama", "hf", "simulated"],
-            help="LLM backend: openai, ollama, or hf (default: openai)",
+            help="LLM backend to use (default: openai)",
         )
         llm_group.add_argument(
             "--llm-model",
             type=str,
             default=None,
-            help="Model name (default: gpt-4o) e.g., gpt-4o-mini, llama3.2:1b, Qwen/Qwen2.5-1.5B-Instruct",
+            help="LLM model name (default: gpt-4o for openai, llama3.2:1b for ollama)",
         )
         llm_group.add_argument(
             "--llm-host",
@@ -215,11 +214,6 @@ class BaseRAGExample(ABC):
         builder.build_index(index_path)
         print(f"Index saved to: {index_path}")
 
-        # Register project directory so leann list can discover this index
-        # The index is saved as args.index_dir/index_name.leann
-        # We want to register the current working directory where the app is run
-        register_project_directory(Path.cwd())
-
         return index_path
 
     async def run_interactive_chat(self, args, index_path: str):

benchmarks/benchmark_no_recompute.py

@@ -1,148 +0,0 @@
-import argparse
-import os
-import time
-from pathlib import Path
-
-from leann import LeannBuilder, LeannSearcher
-
-
-def _meta_exists(index_path: str) -> bool:
-    p = Path(index_path)
-    return (p.parent / f"{p.stem}.meta.json").exists()
-
-
-def ensure_index(index_path: str, backend_name: str, num_docs: int, is_recompute: bool) -> None:
-    # if _meta_exists(index_path):
-    #     return
-    kwargs = {}
-    if backend_name == "hnsw":
-        kwargs["is_compact"] = is_recompute
-    builder = LeannBuilder(
-        backend_name=backend_name,
-        embedding_model=os.getenv("LEANN_EMBED_MODEL", "facebook/contriever"),
-        embedding_mode=os.getenv("LEANN_EMBED_MODE", "sentence-transformers"),
-        graph_degree=32,
-        complexity=64,
-        is_recompute=is_recompute,
-        num_threads=4,
-        **kwargs,
-    )
-    for i in range(num_docs):
-        builder.add_text(
-            f"This is a test document number {i}. It contains some repeated text for benchmarking."
-        )
-    builder.build_index(index_path)
-
-
-def _bench_group(
-    index_path: str,
-    recompute: bool,
-    query: str,
-    repeats: int,
-    complexity: int = 32,
-    top_k: int = 10,
-) -> float:
-    # Independent searcher per group; fixed port when recompute
-    searcher = LeannSearcher(index_path=index_path)
-
-    # Warm-up once
-    _ = searcher.search(
-        query,
-        top_k=top_k,
-        complexity=complexity,
-        recompute_embeddings=recompute,
-    )
-
-    def _once() -> float:
-        t0 = time.time()
-        _ = searcher.search(
-            query,
-            top_k=top_k,
-            complexity=complexity,
-            recompute_embeddings=recompute,
-        )
-        return time.time() - t0
-
-    if repeats <= 1:
-        t = _once()
-    else:
-        vals = [_once() for _ in range(repeats)]
-        vals.sort()
-        t = vals[len(vals) // 2]
-
-    searcher.cleanup()
-    return t
-
-
-def main():
-    parser = argparse.ArgumentParser()
-    parser.add_argument("--num-docs", type=int, default=5000)
-    parser.add_argument("--repeats", type=int, default=3)
-    parser.add_argument("--complexity", type=int, default=32)
-    args = parser.parse_args()
-
-    base = Path.cwd() / ".leann" / "indexes" / f"bench_n{args.num_docs}"
-    base.parent.mkdir(parents=True, exist_ok=True)
-    # ---------- Build HNSW variants ----------
-    hnsw_r = str(base / f"hnsw_recompute_n{args.num_docs}.leann")
-    hnsw_nr = str(base / f"hnsw_norecompute_n{args.num_docs}.leann")
-    ensure_index(hnsw_r, "hnsw", args.num_docs, True)
-    ensure_index(hnsw_nr, "hnsw", args.num_docs, False)
-
-    # ---------- Build DiskANN variants ----------
-    diskann_r = str(base / "diskann_r.leann")
-    diskann_nr = str(base / "diskann_nr.leann")
-    ensure_index(diskann_r, "diskann", args.num_docs, True)
-    ensure_index(diskann_nr, "diskann", args.num_docs, False)
-
-    # ---------- Helpers ----------
-    def _size_for(prefix: str) -> int:
-        p = Path(prefix)
-        base_dir = p.parent
-        stem = p.stem
-        total = 0
-        for f in base_dir.iterdir():
-            if f.is_file() and f.name.startswith(stem):
-                total += f.stat().st_size
-        return total
-
-    # ---------- HNSW benchmark ----------
-    t_hnsw_r = _bench_group(
-        hnsw_r, True, "test document number 42", repeats=args.repeats, complexity=args.complexity
-    )
-    t_hnsw_nr = _bench_group(
-        hnsw_nr, False, "test document number 42", repeats=args.repeats, complexity=args.complexity
-    )
-    size_hnsw_r = _size_for(hnsw_r)
-    size_hnsw_nr = _size_for(hnsw_nr)
-
-    print("Benchmark results (HNSW):")
-    print(f"  recompute=True:  search_time={t_hnsw_r:.3f}s, size={size_hnsw_r / 1024 / 1024:.1f}MB")
-    print(
-        f"  recompute=False: search_time={t_hnsw_nr:.3f}s, size={size_hnsw_nr / 1024 / 1024:.1f}MB"
-    )
-    print("  Expectation: no-recompute should be faster but larger on disk.")
-
-    # ---------- DiskANN benchmark ----------
-    t_diskann_r = _bench_group(
-        diskann_r, True, "DiskANN R test doc 123", repeats=args.repeats, complexity=args.complexity
-    )
-    t_diskann_nr = _bench_group(
-        diskann_nr,
-        False,
-        "DiskANN NR test doc 123",
-        repeats=args.repeats,
-        complexity=args.complexity,
-    )
-    size_diskann_r = _size_for(diskann_r)
-    size_diskann_nr = _size_for(diskann_nr)
-
-    print("\nBenchmark results (DiskANN):")
-    print(f"  build(recompute=True, partition): size={size_diskann_r / 1024 / 1024:.1f}MB")
-    print(f"  build(recompute=False):          size={size_diskann_nr / 1024 / 1024:.1f}MB")
-    print(f"  search recompute=True (final rerank): {t_diskann_r:.3f}s")
-    print(f"  search recompute=False (PQ only):     {t_diskann_nr:.3f}s")
-
-
-if __name__ == "__main__":
-    main()

benchmarks/diskann_vs_hnsw_speed_comparison.py

@@ -10,7 +10,6 @@ This benchmark compares search performance between DiskANN and HNSW backends:
 """
 
 import gc
-import multiprocessing as mp
 import tempfile
 import time
 from pathlib import Path
@@ -18,12 +17,6 @@ from typing import Any
 
 import numpy as np
 
-# Prefer 'fork' start method to avoid POSIX semaphore leaks on macOS
-try:
-    mp.set_start_method("fork", force=True)
-except Exception:
-    pass
-
 
 def create_test_texts(n_docs: int) -> list[str]:
     """Create synthetic test documents for benchmarking."""
@@ -120,10 +113,10 @@ def benchmark_backend(
         ]
         score_validity_rate = len(valid_scores) / len(all_scores) if all_scores else 0
 
-        # Clean up (ensure embedding server shutdown and object GC)
+        # Clean up
         try:
-            if hasattr(searcher, "cleanup"):
-                searcher.cleanup()
+            if hasattr(searcher, "__del__"):
+                searcher.__del__()
             del searcher
             del builder
             gc.collect()
@@ -266,21 +259,10 @@ if __name__ == "__main__":
         print(f"\n❌ Benchmark failed: {e}")
         sys.exit(1)
     finally:
-        # Ensure clean exit (forceful to prevent rare hangs from atexit/threads)
+        # Ensure clean exit
         try:
             gc.collect()
             print("\n🧹 Cleanup completed")
-            # Flush stdio to ensure message is visible before hard-exit
-            try:
-                import sys as _sys
-
-                _sys.stdout.flush()
-                _sys.stderr.flush()
-            except Exception:
-                pass
         except Exception:
             pass
-        # Use os._exit to bypass atexit handlers that may hang in rare cases
-        import os as _os
-
-        _os._exit(0)
+        sys.exit(0)

benchmarks/simple_mac_tpt_test.py

@@ -183,9 +183,6 @@ class Benchmark:
         start_time = time.time()
         with torch.no_grad():
             self.model(input_ids=input_ids, attention_mask=attention_mask)
-        # mps sync
-        if torch.backends.mps.is_available():
-            torch.mps.synchronize()
         end_time = time.time()
 
         return end_time - start_time

docs/configuration-guide.md

@@ -52,7 +52,7 @@ Based on our experience developing LEANN, embedding models fall into three categ
 ### Quick Start: Cloud and Local Embedding Options
 
 **OpenAI Embeddings (Fastest Setup)**
-For immediate testing without local model downloads(also if you [do not have GPU](https://github.com/yichuan-w/LEANN/issues/43) and do not care that much about your document leak, you should use this, we compute the embedding and recompute using openai API):
+For immediate testing without local model downloads:
 ```bash
 # Set OpenAI embeddings (requires OPENAI_API_KEY)
 --embedding-mode openai --embedding-model text-embedding-3-small
@@ -97,23 +97,29 @@ ollama pull nomic-embed-text
 ```
 
 ### DiskANN
-**Best for**: Large datasets, especially when you want `recompute=True`.
+**Best for**: Performance-critical applications and large datasets - **Production-ready with automatic graph partitioning**
 
-**Key advantages:**
-- **Faster search** on large datasets (3x+ speedup vs HNSW in many cases)
-- **Smart storage**: `recompute=True` enables automatic graph partitioning for smaller indexes
-- **Better scaling**: Designed for 100k+ documents
+**How it works:**
+- **Product Quantization (PQ) + Real-time Reranking**: Uses compressed PQ codes for fast graph traversal, then recomputes exact embeddings for final candidates
+- **Automatic Graph Partitioning**: When `is_recompute=True`, automatically partitions large indices and safely removes redundant files to save storage
+- **Superior Speed-Accuracy Trade-off**: Faster search than HNSW while maintaining high accuracy
 
-**Recompute behavior:**
-- `recompute=True` (recommended): Pure PQ traversal + final reranking - faster and enables partitioning
-- `recompute=False`: PQ + partial real distances during traversal - slower but higher accuracy
+**Trade-offs compared to HNSW:**
+- ✅ **Faster search latency** (typically 2-8x speedup)
+- ✅ **Better scaling** for large datasets
+- ✅ **Smart storage management** with automatic partitioning
+- ✅ **Better graph locality** with `--ldg-times` parameter for SSD optimization
+- ⚠️ **Slightly larger index size** due to PQ tables and graph metadata
 
 ```bash
 # Recommended for most use cases
 --backend-name diskann --graph-degree 32 --build-complexity 64
+
+# For large-scale deployments
+--backend-name diskann --graph-degree 64 --build-complexity 128
 ```
 
-**Performance Benchmark**: Run `uv run benchmarks/diskann_vs_hnsw_speed_comparison.py` to compare DiskANN and HNSW on your system.
+**Performance Benchmark**: Run `python benchmarks/diskann_vs_hnsw_speed_comparison.py` to compare DiskANN and HNSW on your system.
 
 ## LLM Selection: Engine and Model Comparison
 
@@ -267,114 +273,24 @@ Every configuration choice involves trade-offs:
 
 The key is finding the right balance for your specific use case. Start small and simple, measure performance, then scale up only where needed.
 
-## Low-resource setups
+## Deep Dive: Critical Configuration Decisions
 
-If you don’t have a local GPU or builds/searches are too slow, use one or more of the options below.
+### When to Disable Recomputation
 
-### 1) Use OpenAI embeddings (no local compute)
-
-Fastest path with zero local GPU requirements. Set your API key and use OpenAI embeddings during build and search:
+LEANN's recomputation feature provides exact distance calculations but can be disabled for extreme QPS requirements:
 
 ```bash
-export OPENAI_API_KEY=sk-...
-
-# Build with OpenAI embeddings
-leann build my-index \
-  --embedding-mode openai \
-  --embedding-model text-embedding-3-small
-
-# Search with OpenAI embeddings (recompute at query time)
-leann search my-index "your query" \
-  --recompute
+--no-recompute  # Disable selective recomputation
 ```
 
-### 2) Run remote builds with SkyPilot (cloud GPU)
-
-Offload embedding generation and index building to a GPU VM using [SkyPilot](https://skypilot.readthedocs.io/en/latest/). A template is provided at `sky/leann-build.yaml`.
-
-```bash
-# One-time: install and configure SkyPilot
-pip install skypilot
-
-# Launch with defaults (L4:1) and mount ./data to ~/leann-data; the build runs automatically
-sky launch -c leann-gpu sky/leann-build.yaml
-
-# Override parameters via -e key=value (optional)
-sky launch -c leann-gpu sky/leann-build.yaml \
-  -e index_name=my-index \
-  -e backend=hnsw \
-  -e embedding_mode=sentence-transformers \
-  -e embedding_model=Qwen/Qwen3-Embedding-0.6B
-
-# Copy the built index back to your local .leann (use rsync)
-rsync -Pavz leann-gpu:~/.leann/indexes/my-index ./.leann/indexes/
-```
-
-### 3) Disable recomputation to trade storage for speed
-
-If you need lower latency and have more storage/memory, disable recomputation. This stores full embeddings and avoids recomputing at search time.
-
-```bash
-# Build without recomputation (HNSW requires non-compact in this mode)
-leann build my-index --no-recompute --no-compact
-
-# Search without recomputation
-leann search my-index "your query" --no-recompute
-```
-
-When to use:
-- Extreme low latency requirements (high QPS, interactive assistants)
-- Read-heavy workloads where storage is cheaper than latency
-- No always-available GPU
-
-Constraints:
-- HNSW: when `--no-recompute` is set, LEANN automatically disables compact mode during build
-- DiskANN: supported; `--no-recompute` skips selective recompute during search
-
-Storage impact:
-- Storing N embeddings of dimension D with float32 requires approximately N × D × 4 bytes
-- Example: 1,000,000 chunks × 768 dims × 4 bytes ≈ 2.86 GB (plus graph/metadata)
-
-Converting an existing index (rebuild required):
-```bash
-# Rebuild in-place (ensure you still have original docs or can regenerate chunks)
-leann build my-index --force --no-recompute --no-compact
-```
-
-Python API usage:
-```python
-from leann import LeannSearcher
-
-searcher = LeannSearcher("/path/to/my-index.leann")
-results = searcher.search("your query", top_k=10, recompute_embeddings=False)
-```
-
-Trade-offs:
-- Lower latency and fewer network hops at query time
-- Significantly higher storage (10–100× vs selective recomputation)
-- Slightly larger memory footprint during build and search
-
-Quick benchmark results (`benchmarks/benchmark_no_recompute.py` with 5k texts, complexity=32):
-
-- HNSW
-
-  ```text
-  recompute=True:  search_time=0.818s, size=1.1MB
-  recompute=False: search_time=0.012s, size=16.6MB
-  ```
-
-- DiskANN
-
-  ```text
-  recompute=True:  search_time=0.041s, size=5.9MB
-  recompute=False: search_time=0.013s, size=24.6MB
-  ```
-
-Conclusion:
-- **HNSW**: `no-recompute` is significantly faster (no embedding recomputation) but requires much more storage (stores all embeddings)
-- **DiskANN**: `no-recompute` uses PQ + partial real distances during traversal (slower but higher accuracy), while `recompute=True` uses pure PQ traversal + final reranking (faster traversal, enables build-time partitioning for smaller storage)
-
+**Trade-offs**:
+- **With recomputation** (default): Exact distances, best quality, higher latency, minimal storage (only stores metadata, recomputes embeddings on-demand)
+- **Without recomputation**: Must store full embeddings, significantly higher memory and storage usage (10-100x more), but faster search
 
+**Disable when**:
+- You have abundant storage and memory
+- Need extremely low latency (< 100ms)
+- Running a read-heavy workload where storage cost is acceptable
 
 ## Further Reading
 

packages/leann-backend-diskann/leann_backend_diskann/diskann_backend.py

@@ -441,14 +441,9 @@ class DiskannSearcher(BaseSearcher):
         else:  # "global"
             use_global_pruning = True
 
-        # Strategy:
-        # - Traversal always uses PQ distances
-        # - If recompute_embeddings=True, do a single final rerank via deferred fetch
-        #   (fetch embeddings for the final candidate set only)
-        # - Do not recompute neighbor distances along the path
-        use_deferred_fetch = True if recompute_embeddings else False
-        recompute_neighors = False  # Expected typo. For backward compatibility.
-
+        # Perform search with suppressed C++ output based on log level
+        use_deferred_fetch = kwargs.get("USE_DEFERRED_FETCH", True)
+        recompute_neighors = False
         with suppress_cpp_output_if_needed():
             labels, distances = self._index.batch_search(
                 query,

packages/leann-backend-diskann/pyproject.toml

@@ -4,8 +4,8 @@ build-backend = "scikit_build_core.build"
 
 [project]
 name = "leann-backend-diskann"
-version = "0.3.1"
-dependencies = ["leann-core==0.3.1", "numpy", "protobuf>=3.19.0"]
+version = "0.2.8"
+dependencies = ["leann-core==0.2.8", "numpy", "protobuf>=3.19.0"]
 
 [tool.scikit-build]
 # Key: simplified CMake path

packages/leann-backend-hnsw/leann_backend_hnsw/hnsw_backend.py

@@ -54,13 +54,12 @@ class HNSWBuilder(LeannBackendBuilderInterface):
         self.efConstruction = self.build_params.setdefault("efConstruction", 200)
         self.distance_metric = self.build_params.setdefault("distance_metric", "mips")
         self.dimensions = self.build_params.get("dimensions")
-        if not self.is_recompute and self.is_compact:
-            # Auto-correct: non-recompute requires non-compact storage for HNSW
-            logger.warning(
-                "is_recompute=False requires non-compact HNSW. Forcing is_compact=False."
-            )
-            self.is_compact = False
-            self.build_params["is_compact"] = False
+        if not self.is_recompute:
+            if self.is_compact:
+                # TODO: support this case @andy
+                raise ValueError(
+                    "is_recompute is False, but is_compact is True. This is not compatible now. change is compact to False and you can use the original HNSW index."
+                )
 
     def build(self, data: np.ndarray, ids: list[str], index_path: str, **kwargs):
         from . import faiss  # type: ignore
@@ -185,11 +184,9 @@ class HNSWSearcher(BaseSearcher):
         """
         from . import faiss  # type: ignore
 
-        if not recompute_embeddings and self.is_pruned:
-            raise RuntimeError(
-                "Recompute is required for pruned/compact HNSW index. "
-                "Re-run search with --recompute, or rebuild with --no-recompute and --no-compact."
-            )
+        if not recompute_embeddings:
+            if self.is_pruned:
+                raise RuntimeError("Recompute is required for pruned index.")
         if recompute_embeddings:
             if zmq_port is None:
                 raise ValueError("zmq_port must be provided if recompute_embeddings is True")

packages/leann-backend-hnsw/pyproject.toml

@@ -6,10 +6,10 @@ build-backend = "scikit_build_core.build"
 
 [project]
 name = "leann-backend-hnsw"
-version = "0.3.1"
+version = "0.2.8"
 description = "Custom-built HNSW (Faiss) backend for the Leann toolkit."
 dependencies = [
-    "leann-core==0.3.1",
+    "leann-core==0.2.8",
     "numpy",
     "pyzmq>=23.0.0",
     "msgpack>=1.0.0",

packages/leann-core/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "leann-core"
-version = "0.3.1"
+version = "0.2.8"
 description = "Core API and plugin system for LEANN"
 readme = "README.md"
 requires-python = ">=3.9"

packages/leann-core/src/leann/api.py

@@ -46,7 +46,6 @@ def compute_embeddings(
             - "sentence-transformers": Use sentence-transformers library (default)
             - "mlx": Use MLX backend for Apple Silicon
             - "openai": Use OpenAI embedding API
-            - "gemini": Use Google Gemini embedding API
         use_server: Whether to use embedding server (True for search, False for build)
 
     Returns:
@@ -205,18 +204,6 @@ class LeannBuilder:
         **backend_kwargs,
     ):
         self.backend_name = backend_name
-        # Normalize incompatible combinations early (for consistent metadata)
-        if backend_name == "hnsw":
-            is_recompute = backend_kwargs.get("is_recompute", True)
-            is_compact = backend_kwargs.get("is_compact", True)
-            if is_recompute is False and is_compact is True:
-                warnings.warn(
-                    "HNSW with is_recompute=False requires non-compact storage. Forcing is_compact=False.",
-                    UserWarning,
-                    stacklevel=2,
-                )
-                backend_kwargs["is_compact"] = False
-
         backend_factory: Optional[LeannBackendFactoryInterface] = BACKEND_REGISTRY.get(backend_name)
         if backend_factory is None:
             raise ValueError(f"Backend '{backend_name}' not found or not registered.")
@@ -307,23 +294,6 @@ class LeannBuilder:
     def build_index(self, index_path: str):
         if not self.chunks:
             raise ValueError("No chunks added.")
-
-        # Filter out invalid/empty text chunks early to keep passage and embedding counts aligned
-        valid_chunks: list[dict[str, Any]] = []
-        skipped = 0
-        for chunk in self.chunks:
-            text = chunk.get("text", "")
-            if isinstance(text, str) and text.strip():
-                valid_chunks.append(chunk)
-            else:
-                skipped += 1
-        if skipped > 0:
-            print(
-                f"Warning: Skipping {skipped} empty/invalid text chunk(s). Processing {len(valid_chunks)} valid chunks"
-            )
-            self.chunks = valid_chunks
-            if not self.chunks:
-                raise ValueError("All provided chunks are empty or invalid. Nothing to index.")
         if self.dimensions is None:
             self.dimensions = len(
                 compute_embeddings(
@@ -553,7 +523,6 @@ class LeannSearcher:
         self.embedding_model = self.meta_data["embedding_model"]
         # Support both old and new format
         self.embedding_mode = self.meta_data.get("embedding_mode", "sentence-transformers")
-        # Delegate portability handling to PassageManager
         self.passage_manager = PassageManager(
             self.meta_data.get("passage_sources", []), metadata_file_path=self.meta_path_str
         )
@@ -614,7 +583,7 @@ class LeannSearcher:
             zmq_port=zmq_port,
         )
         # logger.info(f"  Generated embedding shape: {query_embedding.shape}")
-        # time.time() - start_time
+        time.time() - start_time
         # logger.info(f"  Embedding time: {embedding_time} seconds")
 
         start_time = time.time()
@@ -680,26 +649,8 @@ class LeannSearcher:
         This method should be called after you're done using the searcher,
         especially in test environments or batch processing scenarios.
         """
-        backend = getattr(self.backend_impl, "embedding_server_manager", None)
-        if backend is not None:
-            backend.stop_server()
-
-    # Enable automatic cleanup patterns
-    def __enter__(self):
-        return self
-
-    def __exit__(self, exc_type, exc, tb):
-        try:
-            self.cleanup()
-        except Exception:
-            pass
-
-    def __del__(self):
-        try:
-            self.cleanup()
-        except Exception:
-            # Avoid noisy errors during interpreter shutdown
-            pass
+        if hasattr(self.backend_impl, "embedding_server_manager"):
+            self.backend_impl.embedding_server_manager.stop_server()
 
 
 class LeannChat:
@@ -779,19 +730,3 @@ class LeannChat:
         """
         if hasattr(self.searcher, "cleanup"):
             self.searcher.cleanup()
-
-    # Enable automatic cleanup patterns
-    def __enter__(self):
-        return self
-
-    def __exit__(self, exc_type, exc, tb):
-        try:
-            self.cleanup()
-        except Exception:
-            pass
-
-    def __del__(self):
-        try:
-            self.cleanup()
-        except Exception:
-            pass

packages/leann-core/src/leann/chat.py

@@ -422,6 +422,7 @@ class LLMInterface(ABC):
                 top_k=10,
                 complexity=64,
                 beam_width=8,
+                USE_DEFERRED_FETCH=True,
                 skip_search_reorder=True,
                 recompute_beighbor_embeddings=True,
                 dedup_node_dis=True,
@@ -433,6 +434,7 @@ class LLMInterface(ABC):
         Supported kwargs:
             - complexity (int): Search complexity parameter (default: 32)
             - beam_width (int): Beam width for search (default: 4)
+            - USE_DEFERRED_FETCH (bool): Enable deferred fetch mode (default: False)
             - skip_search_reorder (bool): Skip search reorder step (default: False)
             - recompute_beighbor_embeddings (bool): Enable ZMQ embedding server for neighbor recomputation (default: False)
             - dedup_node_dis (bool): Deduplicate nodes by distance (default: False)
@@ -680,60 +682,6 @@ class HFChat(LLMInterface):
         return response.strip()
 
 
-class GeminiChat(LLMInterface):
-    """LLM interface for Google Gemini models."""
-
-    def __init__(self, model: str = "gemini-2.5-flash", api_key: Optional[str] = None):
-        self.model = model
-        self.api_key = api_key or os.getenv("GEMINI_API_KEY")
-
-        if not self.api_key:
-            raise ValueError(
-                "Gemini API key is required. Set GEMINI_API_KEY environment variable or pass api_key parameter."
-            )
-
-        logger.info(f"Initializing Gemini Chat with model='{model}'")
-
-        try:
-            import google.genai as genai
-
-            self.client = genai.Client(api_key=self.api_key)
-        except ImportError:
-            raise ImportError(
-                "The 'google-genai' library is required for Gemini models. Please install it with 'uv pip install google-genai'."
-            )
-
-    def ask(self, prompt: str, **kwargs) -> str:
-        logger.info(f"Sending request to Gemini with model {self.model}")
-
-        try:
-            from google.genai.types import GenerateContentConfig
-
-            generation_config = GenerateContentConfig(
-                temperature=kwargs.get("temperature", 0.7),
-                max_output_tokens=kwargs.get("max_tokens", 1000),
-            )
-
-            # Handle top_p parameter
-            if "top_p" in kwargs:
-                generation_config.top_p = kwargs["top_p"]
-
-            response = self.client.models.generate_content(
-                model=self.model,
-                contents=prompt,
-                config=generation_config,
-            )
-            # Handle potential None response text
-            response_text = response.text
-            if response_text is None:
-                logger.warning("Gemini returned None response text")
-                return ""
-            return response_text.strip()
-        except Exception as e:
-            logger.error(f"Error communicating with Gemini: {e}")
-            return f"Error: Could not get a response from Gemini. Details: {e}"
-
-
 class OpenAIChat(LLMInterface):
     """LLM interface for OpenAI models."""
 
@@ -847,8 +795,6 @@ def get_llm(llm_config: Optional[dict[str, Any]] = None) -> LLMInterface:
         return HFChat(model_name=model or "deepseek-ai/deepseek-llm-7b-chat")
     elif llm_type == "openai":
         return OpenAIChat(model=model or "gpt-4o", api_key=llm_config.get("api_key"))
-    elif llm_type == "gemini":
-        return GeminiChat(model=model or "gemini-2.5-flash", api_key=llm_config.get("api_key"))
     elif llm_type == "simulated":
         return SimulatedChat()
     else:

packages/leann-core/src/leann/cli.py

@@ -1,14 +1,13 @@
 import argparse
 import asyncio
 from pathlib import Path
-from typing import Optional, Union
+from typing import Union
 
 from llama_index.core import SimpleDirectoryReader
 from llama_index.core.node_parser import SentenceSplitter
 from tqdm import tqdm
 
 from .api import LeannBuilder, LeannChat, LeannSearcher
-from .registry import register_project_directory
 
 
 def extract_pdf_text_with_pymupdf(file_path: str) -> str:
@@ -73,7 +72,7 @@ class LeannCLI:
     def create_parser(self) -> argparse.ArgumentParser:
         parser = argparse.ArgumentParser(
             prog="leann",
-            description="The smallest vector index in the world. RAG Everything with LEANN!",
+            description="LEANN - Local Enhanced AI Navigation",
             formatter_class=argparse.RawDescriptionHelpFormatter,
             epilog="""
 Examples:
@@ -85,7 +84,6 @@ Examples:
   leann search my-docs "query"                                           # Search in my-docs index
   leann ask my-docs "question"                                           # Ask my-docs index
   leann list                                                             # List all stored indexes
-  leann remove my-docs                                                   # Remove an index (local first, then global)
             """,
         )
 
@@ -104,18 +102,9 @@ Examples:
             help="Documents directories and/or files (default: current directory)",
         )
         build_parser.add_argument(
-            "--backend",
-            type=str,
-            default="hnsw",
-            choices=["hnsw", "diskann"],
-            help="Backend to use (default: hnsw)",
-        )
-        build_parser.add_argument(
-            "--embedding-model",
-            type=str,
-            default="facebook/contriever",
-            help="Embedding model (default: facebook/contriever)",
+            "--backend", type=str, default="hnsw", choices=["hnsw", "diskann"]
         )
+        build_parser.add_argument("--embedding-model", type=str, default="facebook/contriever")
         build_parser.add_argument(
             "--embedding-mode",
             type=str,
@@ -123,88 +112,36 @@ Examples:
             choices=["sentence-transformers", "openai", "mlx", "ollama"],
             help="Embedding backend mode (default: sentence-transformers)",
         )
-        build_parser.add_argument(
-            "--force", "-f", action="store_true", help="Force rebuild existing index"
-        )
-        build_parser.add_argument(
-            "--graph-degree", type=int, default=32, help="Graph degree (default: 32)"
-        )
-        build_parser.add_argument(
-            "--complexity", type=int, default=64, help="Build complexity (default: 64)"
-        )
+        build_parser.add_argument("--force", "-f", action="store_true", help="Force rebuild")
+        build_parser.add_argument("--graph-degree", type=int, default=32)
+        build_parser.add_argument("--complexity", type=int, default=64)
         build_parser.add_argument("--num-threads", type=int, default=1)
-        build_parser.add_argument(
-            "--compact",
-            action=argparse.BooleanOptionalAction,
-            default=True,
-            help="Use compact storage (default: true). Must be `no-compact` for `no-recompute` build.",
-        )
-        build_parser.add_argument(
-            "--recompute",
-            action=argparse.BooleanOptionalAction,
-            default=True,
-            help="Enable recomputation (default: true)",
-        )
+        build_parser.add_argument("--compact", action="store_true", default=True)
+        build_parser.add_argument("--recompute", action="store_true", default=True)
         build_parser.add_argument(
             "--file-types",
             type=str,
             help="Comma-separated list of file extensions to include (e.g., '.txt,.pdf,.pptx'). If not specified, uses default supported types.",
         )
-        build_parser.add_argument(
-            "--include-hidden",
-            action=argparse.BooleanOptionalAction,
-            default=False,
-            help="Include hidden files and directories (paths starting with '.') during indexing (default: false)",
-        )
-        build_parser.add_argument(
-            "--doc-chunk-size",
-            type=int,
-            default=256,
-            help="Document chunk size in tokens/characters (default: 256)",
-        )
-        build_parser.add_argument(
-            "--doc-chunk-overlap",
-            type=int,
-            default=128,
-            help="Document chunk overlap (default: 128)",
-        )
-        build_parser.add_argument(
-            "--code-chunk-size",
-            type=int,
-            default=512,
-            help="Code chunk size in tokens/lines (default: 512)",
-        )
-        build_parser.add_argument(
-            "--code-chunk-overlap",
-            type=int,
-            default=50,
-            help="Code chunk overlap (default: 50)",
-        )
 
         # Search command
         search_parser = subparsers.add_parser("search", help="Search documents")
         search_parser.add_argument("index_name", help="Index name")
         search_parser.add_argument("query", help="Search query")
-        search_parser.add_argument(
-            "--top-k", type=int, default=5, help="Number of results (default: 5)"
-        )
-        search_parser.add_argument(
-            "--complexity", type=int, default=64, help="Search complexity (default: 64)"
-        )
+        search_parser.add_argument("--top-k", type=int, default=5)
+        search_parser.add_argument("--complexity", type=int, default=64)
         search_parser.add_argument("--beam-width", type=int, default=1)
         search_parser.add_argument("--prune-ratio", type=float, default=0.0)
         search_parser.add_argument(
-            "--recompute",
-            dest="recompute_embeddings",
-            action=argparse.BooleanOptionalAction,
+            "--recompute-embeddings",
+            action="store_true",
             default=True,
-            help="Enable/disable embedding recomputation (default: enabled). Should not do a `no-recompute` search in a `recompute` build.",
+            help="Recompute embeddings (default: True)",
         )
         search_parser.add_argument(
             "--pruning-strategy",
             choices=["global", "local", "proportional"],
             default="global",
-            help="Pruning strategy (default: global)",
         )
 
         # Ask command
@@ -215,27 +152,19 @@ Examples:
             type=str,
             default="ollama",
             choices=["simulated", "ollama", "hf", "openai"],
-            help="LLM provider (default: ollama)",
-        )
-        ask_parser.add_argument(
-            "--model", type=str, default="qwen3:8b", help="Model name (default: qwen3:8b)"
         )
+        ask_parser.add_argument("--model", type=str, default="qwen3:8b")
         ask_parser.add_argument("--host", type=str, default="http://localhost:11434")
-        ask_parser.add_argument(
-            "--interactive", "-i", action="store_true", help="Interactive chat mode"
-        )
-        ask_parser.add_argument(
-            "--top-k", type=int, default=20, help="Retrieval count (default: 20)"
-        )
+        ask_parser.add_argument("--interactive", "-i", action="store_true")
+        ask_parser.add_argument("--top-k", type=int, default=20)
         ask_parser.add_argument("--complexity", type=int, default=32)
         ask_parser.add_argument("--beam-width", type=int, default=1)
         ask_parser.add_argument("--prune-ratio", type=float, default=0.0)
         ask_parser.add_argument(
-            "--recompute",
-            dest="recompute_embeddings",
-            action=argparse.BooleanOptionalAction,
+            "--recompute-embeddings",
+            action="store_true",
             default=True,
-            help="Enable/disable embedding recomputation during ask (default: enabled)",
+            help="Recompute embeddings (default: True)",
         )
         ask_parser.add_argument(
             "--pruning-strategy",
@@ -253,18 +182,35 @@ Examples:
         # List command
         subparsers.add_parser("list", help="List all indexes")
 
-        # Remove command
-        remove_parser = subparsers.add_parser("remove", help="Remove an index")
-        remove_parser.add_argument("index_name", help="Index name to remove")
-        remove_parser.add_argument(
-            "--force", "-f", action="store_true", help="Force removal without confirmation"
-        )
-
         return parser
 
     def register_project_dir(self):
         """Register current project directory in global registry"""
-        register_project_directory()
+        global_registry = Path.home() / ".leann" / "projects.json"
+        global_registry.parent.mkdir(exist_ok=True)
+
+        current_dir = str(Path.cwd())
+
+        # Load existing registry
+        projects = []
+        if global_registry.exists():
+            try:
+                import json
+
+                with open(global_registry) as f:
+                    projects = json.load(f)
+            except Exception:
+                projects = []
+
+        # Add current directory if not already present
+        if current_dir not in projects:
+            projects.append(current_dir)
+
+        # Save registry
+        import json
+
+        with open(global_registry, "w") as f:
+            json.dump(projects, f, indent=2)
 
     def _build_gitignore_parser(self, docs_dir: str):
         """Build gitignore parser using gitignore-parser library."""
@@ -324,6 +270,8 @@ Examples:
             return False
 
     def list_indexes(self):
+        print("Stored LEANN indexes:")
+
         # Get all project directories with .leann
         global_registry = Path.home() / ".leann" / "projects.json"
         all_projects = []
@@ -349,485 +297,58 @@ Examples:
         if (current_path / ".leann" / "indexes").exists() and current_path not in valid_projects:
             valid_projects.append(current_path)
 
-        # Separate current and other projects
-        other_projects = []
-
-        for project_path in valid_projects:
-            if project_path != current_path:
-                other_projects.append(project_path)
-
-        print("📚 LEANN Indexes")
-        print("=" * 50)
+        if not valid_projects:
+            print(
+                "No indexes found. Use 'leann build <name> --docs <dir> [<dir2> ...]' to create one."
+            )
+            return
 
         total_indexes = 0
-        current_indexes_count = 0
+        current_dir = Path.cwd()
 
-        # Show current project first (most important)
-        print("\n🏠 Current Project")
-        print(f"   {current_path}")
-        print("   " + "─" * 45)
-
-        current_indexes = self._discover_indexes_in_project(current_path)
-        if current_indexes:
-            for idx in current_indexes:
-                total_indexes += 1
-                current_indexes_count += 1
-                type_icon = "📁" if idx["type"] == "cli" else "📄"
-                print(f"   {current_indexes_count}. {type_icon} {idx['name']} {idx['status']}")
-                if idx["size_mb"] > 0:
-                    print(f"      📦 Size: {idx['size_mb']:.1f} MB")
-        else:
-            print("   📭 No indexes in current project")
-
-        # Show other projects (reference information)
-        if other_projects:
-            print("\n\n🗂️  Other Projects")
-            print("   " + "─" * 45)
-
-            for project_path in other_projects:
-                project_indexes = self._discover_indexes_in_project(project_path)
-                if not project_indexes:
-                    continue
-
-                print(f"\n   📂 {project_path.name}")
-                print(f"      {project_path}")
-
-                for idx in project_indexes:
-                    total_indexes += 1
-                    type_icon = "📁" if idx["type"] == "cli" else "📄"
-                    print(f"      • {type_icon} {idx['name']} {idx['status']}")
-                    if idx["size_mb"] > 0:
-                        print(f"        📦 {idx['size_mb']:.1f} MB")
-
-        # Summary and usage info
-        print("\n" + "=" * 50)
-        if total_indexes == 0:
-            print("💡 Get started:")
-            print("   leann build my-docs --docs ./documents")
-        else:
-            # Count only projects that have at least one discoverable index
-            projects_count = sum(
-                1 for p in valid_projects if len(self._discover_indexes_in_project(p)) > 0
-            )
-            print(f"📊 Total: {total_indexes} indexes across {projects_count} projects")
-
-            if current_indexes_count > 0:
-                print("\n💫 Quick start (current project):")
-                # Get first index from current project for example
-                current_indexes_dir = current_path / ".leann" / "indexes"
-                if current_indexes_dir.exists():
-                    current_index_dirs = [d for d in current_indexes_dir.iterdir() if d.is_dir()]
-                    if current_index_dirs:
-                        example_name = current_index_dirs[0].name
-                        print(f'   leann search {example_name} "your query"')
-                        print(f"   leann ask {example_name} --interactive")
-            else:
-                print("\n💡 Create your first index:")
-                print("   leann build my-docs --docs ./documents")
-
-    def _discover_indexes_in_project(self, project_path: Path):
-        """Discover all indexes in a project directory (both CLI and apps formats)"""
-        indexes = []
-
-        # 1. CLI format: .leann/indexes/index_name/
-        cli_indexes_dir = project_path / ".leann" / "indexes"
-        if cli_indexes_dir.exists():
-            for index_dir in cli_indexes_dir.iterdir():
-                if index_dir.is_dir():
-                    meta_file = index_dir / "documents.leann.meta.json"
-                    status = "✅" if meta_file.exists() else "❌"
-
-                    size_mb = 0
-                    if meta_file.exists():
-                        try:
-                            size_mb = sum(
-                                f.stat().st_size for f in index_dir.iterdir() if f.is_file()
-                            ) / (1024 * 1024)
-                        except (OSError, PermissionError):
-                            pass
-
-                    indexes.append(
-                        {
-                            "name": index_dir.name,
-                            "type": "cli",
-                            "status": status,
-                            "size_mb": size_mb,
-                            "path": index_dir,
-                        }
-                    )
-
-        # 2. Apps format: *.leann.meta.json files anywhere in the project
-        cli_indexes_dir = project_path / ".leann" / "indexes"
-        for meta_file in project_path.rglob("*.leann.meta.json"):
-            if meta_file.is_file():
-                # Skip CLI-built indexes (which store meta under .leann/indexes/<name>/)
-                try:
-                    if cli_indexes_dir.exists() and cli_indexes_dir in meta_file.parents:
-                        continue
-                except Exception:
-                    pass
-                # Use the parent directory name as the app index display name
-                display_name = meta_file.parent.name
-                # Extract file base used to store files
-                file_base = meta_file.name.replace(".leann.meta.json", "")
-
-                # Apps indexes are considered complete if the .leann.meta.json file exists
-                status = "✅"
-
-                # Calculate total size of all related files (use file base)
-                size_mb = 0
-                try:
-                    index_dir = meta_file.parent
-                    for related_file in index_dir.glob(f"{file_base}.leann*"):
-                        size_mb += related_file.stat().st_size / (1024 * 1024)
-                except (OSError, PermissionError):
-                    pass
-
-                indexes.append(
-                    {
-                        "name": display_name,
-                        "type": "app",
-                        "status": status,
-                        "size_mb": size_mb,
-                        "path": meta_file,
-                    }
-                )
-
-        return indexes
-
-    def remove_index(self, index_name: str, force: bool = False):
-        """Safely remove an index - always show all matches for transparency"""
-
-        # Always do a comprehensive search for safety
-        print(f"🔍 Searching for all indexes named '{index_name}'...")
-        all_matches = self._find_all_matching_indexes(index_name)
-
-        if not all_matches:
-            print(f"❌ Index '{index_name}' not found in any project.")
-            return False
-
-        if len(all_matches) == 1:
-            return self._remove_single_match(all_matches[0], index_name, force)
-        else:
-            return self._remove_from_multiple_matches(all_matches, index_name, force)
-
-    def _find_all_matching_indexes(self, index_name: str):
-        """Find all indexes with the given name across all projects"""
-        matches = []
-
-        # Get all registered projects
-        global_registry = Path.home() / ".leann" / "projects.json"
-        all_projects = []
-
-        if global_registry.exists():
-            try:
-                import json
-
-                with open(global_registry) as f:
-                    all_projects = json.load(f)
-            except Exception:
-                pass
-
-        # Always include current project
-        current_path = Path.cwd()
-        if str(current_path) not in all_projects:
-            all_projects.append(str(current_path))
-
-        # Search across all projects
-        for project_dir in all_projects:
-            project_path = Path(project_dir)
-            if not project_path.exists():
+        for project_path in valid_projects:
+            indexes_dir = project_path / ".leann" / "indexes"
+            if not indexes_dir.exists():
                 continue
 
-            # 1) CLI-format index under .leann/indexes/<name>
-            index_dir = project_path / ".leann" / "indexes" / index_name
-            if index_dir.exists():
-                is_current = project_path == current_path
-                matches.append(
-                    {
-                        "project_path": project_path,
-                        "index_dir": index_dir,
-                        "is_current": is_current,
-                        "kind": "cli",
-                    }
-                )
+            index_dirs = [d for d in indexes_dir.iterdir() if d.is_dir()]
+            if not index_dirs:
+                continue
 
-            # 2) App-format indexes
-            # We support two ways of addressing apps:
-            #   a) by the file base (e.g., `pdf_documents`)
-            #   b) by the parent directory name (e.g., `new_txt`)
-            seen_app_meta = set()
-
-            # 2a) by file base
-            for meta_file in project_path.rglob(f"{index_name}.leann.meta.json"):
-                if meta_file.is_file():
-                    # Skip CLI-built indexes' meta under .leann/indexes
-                    try:
-                        cli_indexes_dir = project_path / ".leann" / "indexes"
-                        if cli_indexes_dir.exists() and cli_indexes_dir in meta_file.parents:
-                            continue
-                    except Exception:
-                        pass
-                    is_current = project_path == current_path
-                    key = (str(project_path), str(meta_file))
-                    if key in seen_app_meta:
-                        continue
-                    seen_app_meta.add(key)
-                    matches.append(
-                        {
-                            "project_path": project_path,
-                            "files_dir": meta_file.parent,
-                            "meta_file": meta_file,
-                            "is_current": is_current,
-                            "kind": "app",
-                            "display_name": meta_file.parent.name,
-                            "file_base": meta_file.name.replace(".leann.meta.json", ""),
-                        }
-                    )
-
-            # 2b) by parent directory name
-            for meta_file in project_path.rglob("*.leann.meta.json"):
-                if meta_file.is_file() and meta_file.parent.name == index_name:
-                    # Skip CLI-built indexes' meta under .leann/indexes
-                    try:
-                        cli_indexes_dir = project_path / ".leann" / "indexes"
-                        if cli_indexes_dir.exists() and cli_indexes_dir in meta_file.parents:
-                            continue
-                    except Exception:
-                        pass
-                    is_current = project_path == current_path
-                    key = (str(project_path), str(meta_file))
-                    if key in seen_app_meta:
-                        continue
-                    seen_app_meta.add(key)
-                    matches.append(
-                        {
-                            "project_path": project_path,
-                            "files_dir": meta_file.parent,
-                            "meta_file": meta_file,
-                            "is_current": is_current,
-                            "kind": "app",
-                            "display_name": meta_file.parent.name,
-                            "file_base": meta_file.name.replace(".leann.meta.json", ""),
-                        }
-                    )
-
-        # Sort: current project first, then by project name
-        matches.sort(key=lambda x: (not x["is_current"], x["project_path"].name))
-        return matches
-
-    def _remove_single_match(self, match, index_name: str, force: bool):
-        """Handle removal when only one match is found"""
-        project_path = match["project_path"]
-        is_current = match["is_current"]
-        kind = match.get("kind", "cli")
-
-        if is_current:
-            location_info = "current project"
-            emoji = "🏠"
-        else:
-            location_info = f"other project '{project_path.name}'"
-            emoji = "📂"
-
-        print(f"✅ Found 1 index named '{index_name}':")
-        print(f"   {emoji} Location: {location_info}")
-        if kind == "cli":
-            print(f"   📍 Path: {project_path / '.leann' / 'indexes' / index_name}")
-        else:
-            print(f"   📍 Meta: {match['meta_file']}")
-
-        if not force:
-            if not is_current:
-                print("\n⚠️  CROSS-PROJECT REMOVAL!")
-                print("   This will delete the index from another project.")
-
-            response = input(f"   ❓ Confirm removal from {location_info}? (y/N): ").strip().lower()
-            if response not in ["y", "yes"]:
-                print("   ❌ Removal cancelled.")
-                return False
-
-        if kind == "cli":
-            return self._delete_index_directory(
-                match["index_dir"],
-                index_name,
-                project_path if not is_current else None,
-                is_app=False,
-            )
-        else:
-            return self._delete_index_directory(
-                match["files_dir"],
-                match.get("display_name", index_name),
-                project_path if not is_current else None,
-                is_app=True,
-                meta_file=match.get("meta_file"),
-                app_file_base=match.get("file_base"),
-            )
-
-    def _remove_from_multiple_matches(self, matches, index_name: str, force: bool):
-        """Handle removal when multiple matches are found"""
-
-        print(f"⚠️  Found {len(matches)} indexes named '{index_name}':")
-        print("   " + "─" * 50)
-
-        for i, match in enumerate(matches, 1):
-            project_path = match["project_path"]
-            is_current = match["is_current"]
-            kind = match.get("kind", "cli")
-
-            if is_current:
-                print(f"   {i}. 🏠 Current project ({'CLI' if kind == 'cli' else 'APP'})")
+            # Show project header
+            if project_path == current_dir:
+                print(f"\n📁 Current project ({project_path}):")
             else:
-                print(f"   {i}. 📂 {project_path.name} ({'CLI' if kind == 'cli' else 'APP'})")
+                print(f"\n📂 {project_path}:")
 
-            # Show path details
-            if kind == "cli":
-                print(f"      📍 {project_path / '.leann' / 'indexes' / index_name}")
-            else:
-                print(f"      📍 {match['meta_file']}")
+            for index_dir in index_dirs:
+                total_indexes += 1
+                index_name = index_dir.name
+                meta_file = index_dir / "documents.leann.meta.json"
+                status = "✓" if meta_file.exists() else "✗"
 
-            # Show size info
-            try:
-                if kind == "cli":
-                    size_mb = sum(
-                        f.stat().st_size for f in match["index_dir"].iterdir() if f.is_file()
-                    ) / (1024 * 1024)
-                else:
-                    file_base = match.get("file_base")
-                    size_mb = 0.0
-                    if file_base:
-                        size_mb = sum(
-                            f.stat().st_size
-                            for f in match["files_dir"].glob(f"{file_base}.leann*")
-                            if f.is_file()
-                        ) / (1024 * 1024)
-                print(f"      📦 Size: {size_mb:.1f} MB")
-            except (OSError, PermissionError):
-                pass
-
-        print("   " + "─" * 50)
-
-        if force:
-            print("   ❌ Multiple matches found, but --force specified.")
-            print("   Please run without --force to choose which one to remove.")
-            return False
-
-        try:
-            choice = input(
-                f"   ❓ Which one to remove? (1-{len(matches)}, or 'c' to cancel): "
-            ).strip()
-            if choice.lower() == "c":
-                print("   ❌ Removal cancelled.")
-                return False
-
-            choice_idx = int(choice) - 1
-            if 0 <= choice_idx < len(matches):
-                selected_match = matches[choice_idx]
-                project_path = selected_match["project_path"]
-                is_current = selected_match["is_current"]
-                kind = selected_match.get("kind", "cli")
-
-                location = "current project" if is_current else f"'{project_path.name}' project"
-                print(f"   🎯 Selected: Remove from {location}")
-
-                # Final confirmation for safety
-                confirm = input(
-                    f"   ❓ FINAL CONFIRMATION - Type '{index_name}' to proceed: "
-                ).strip()
-                if confirm != index_name:
-                    print("   ❌ Confirmation failed. Removal cancelled.")
-                    return False
-
-                if kind == "cli":
-                    return self._delete_index_directory(
-                        selected_match["index_dir"],
-                        index_name,
-                        project_path if not is_current else None,
-                        is_app=False,
+                print(f"  {total_indexes}. {index_name} [{status}]")
+                if status == "✓":
+                    size_mb = sum(f.stat().st_size for f in index_dir.iterdir() if f.is_file()) / (
+                        1024 * 1024
                     )
-                else:
-                    return self._delete_index_directory(
-                        selected_match["files_dir"],
-                        selected_match.get("display_name", index_name),
-                        project_path if not is_current else None,
-                        is_app=True,
-                        meta_file=selected_match.get("meta_file"),
-                        app_file_base=selected_match.get("file_base"),
-                    )
-            else:
-                print("   ❌ Invalid choice. Removal cancelled.")
-                return False
+                    print(f"     Size: {size_mb:.1f} MB")
 
-        except (ValueError, KeyboardInterrupt):
-            print("\n   ❌ Invalid input. Removal cancelled.")
-            return False
+        if total_indexes > 0:
+            print(f"\nTotal: {total_indexes} indexes across {len(valid_projects)} projects")
+            print("\nUsage (current project only):")
 
-    def _delete_index_directory(
-        self,
-        index_dir: Path,
-        index_display_name: str,
-        project_path: Optional[Path] = None,
-        is_app: bool = False,
-        meta_file: Optional[Path] = None,
-        app_file_base: Optional[str] = None,
-    ):
-        """Delete a CLI index directory or APP index files safely."""
-        try:
-            if is_app:
-                removed = 0
-                errors = 0
-                # Delete only files that belong to this app index (based on file base)
-                pattern_base = app_file_base or ""
-                for f in index_dir.glob(f"{pattern_base}.leann*"):
-                    try:
-                        f.unlink()
-                        removed += 1
-                    except Exception:
-                        errors += 1
-                # Best-effort: also remove the meta file if specified and still exists
-                if meta_file and meta_file.exists():
-                    try:
-                        meta_file.unlink()
-                        removed += 1
-                    except Exception:
-                        errors += 1
-
-                if removed > 0 and errors == 0:
-                    if project_path:
-                        print(
-                            f"✅ App index '{index_display_name}' removed from {project_path.name}"
-                        )
-                    else:
-                        print(f"✅ App index '{index_display_name}' removed successfully")
-                    return True
-                elif removed > 0 and errors > 0:
-                    print(
-                        f"⚠️  App index '{index_display_name}' partially removed (some files couldn't be deleted)"
-                    )
-                    return True
-                else:
-                    print(
-                        f"❌ No files found to remove for app index '{index_display_name}' in {index_dir}"
-                    )
-                    return False
-            else:
-                import shutil
-
-                shutil.rmtree(index_dir)
-
-                if project_path:
-                    print(f"✅ Index '{index_display_name}' removed from {project_path.name}")
-                else:
-                    print(f"✅ Index '{index_display_name}' removed successfully")
-                return True
-        except Exception as e:
-            print(f"❌ Error removing index '{index_display_name}': {e}")
-            return False
+            # Show example from current project
+            current_indexes_dir = current_dir / ".leann" / "indexes"
+            if current_indexes_dir.exists():
+                current_index_dirs = [d for d in current_indexes_dir.iterdir() if d.is_dir()]
+                if current_index_dirs:
+                    example_name = current_index_dirs[0].name
+                    print(f'  leann search {example_name} "your query"')
+                    print(f"  leann ask {example_name} --interactive")
 
     def load_documents(
-        self,
-        docs_paths: Union[str, list],
-        custom_file_types: Union[str, None] = None,
-        include_hidden: bool = False,
+        self, docs_paths: Union[str, list], custom_file_types: Union[str, None] = None
     ):
         # Handle both single path (string) and multiple paths (list) for backward compatibility
         if isinstance(docs_paths, str):
@@ -871,10 +392,6 @@ Examples:
 
         all_documents = []
 
-        # Helper to detect hidden path components
-        def _path_has_hidden_segment(p: Path) -> bool:
-            return any(part.startswith(".") and part not in [".", ".."] for part in p.parts)
-
         # First, process individual files if any
         if files:
             print(f"\n🔄 Processing {len(files)} individual file{'s' if len(files) > 1 else ''}...")
@@ -887,12 +404,8 @@ Examples:
 
                 files_by_dir = defaultdict(list)
                 for file_path in files:
-                    file_path_obj = Path(file_path)
-                    if not include_hidden and _path_has_hidden_segment(file_path_obj):
-                        print(f"  ⚠️  Skipping hidden file: {file_path}")
-                        continue
-                    parent_dir = str(file_path_obj.parent)
-                    files_by_dir[parent_dir].append(str(file_path_obj))
+                    parent_dir = str(Path(file_path).parent)
+                    files_by_dir[parent_dir].append(file_path)
 
                 # Load files from each parent directory
                 for parent_dir, file_list in files_by_dir.items():
@@ -903,7 +416,6 @@ Examples:
                         file_docs = SimpleDirectoryReader(
                             parent_dir,
                             input_files=file_list,
-                            # exclude_hidden only affects directory scans; input_files are explicit
                             filename_as_id=True,
                         ).load_data()
                         all_documents.extend(file_docs)
@@ -1002,8 +514,6 @@ Examples:
                     # Check if file matches any exclude pattern
                     try:
                         relative_path = file_path.relative_to(docs_path)
-                        if not include_hidden and _path_has_hidden_segment(relative_path):
-                            continue
                         if self._should_exclude_file(relative_path, gitignore_matches):
                             continue
                     except ValueError:
@@ -1031,7 +541,6 @@ Examples:
                         try:
                             default_docs = SimpleDirectoryReader(
                                 str(file_path.parent),
-                                exclude_hidden=not include_hidden,
                                 filename_as_id=True,
                                 required_exts=[file_path.suffix],
                             ).load_data()
@@ -1060,7 +569,6 @@ Examples:
                     encoding="utf-8",
                     required_exts=code_extensions,
                     file_extractor={},  # Use default extractors
-                    exclude_hidden=not include_hidden,
                     filename_as_id=True,
                 ).load_data(show_progress=True)
 
@@ -1179,40 +687,7 @@ Examples:
             print(f"Index '{index_name}' already exists. Use --force to rebuild.")
             return
 
-        # Configure chunking based on CLI args before loading documents
-        # Guard against invalid configurations
-        doc_chunk_size = max(1, int(args.doc_chunk_size))
-        doc_chunk_overlap = max(0, int(args.doc_chunk_overlap))
-        if doc_chunk_overlap >= doc_chunk_size:
-            print(
-                f"⚠️  Adjusting doc chunk overlap from {doc_chunk_overlap} to {doc_chunk_size - 1} (must be < chunk size)"
-            )
-            doc_chunk_overlap = doc_chunk_size - 1
-
-        code_chunk_size = max(1, int(args.code_chunk_size))
-        code_chunk_overlap = max(0, int(args.code_chunk_overlap))
-        if code_chunk_overlap >= code_chunk_size:
-            print(
-                f"⚠️  Adjusting code chunk overlap from {code_chunk_overlap} to {code_chunk_size - 1} (must be < chunk size)"
-            )
-            code_chunk_overlap = code_chunk_size - 1
-
-        self.node_parser = SentenceSplitter(
-            chunk_size=doc_chunk_size,
-            chunk_overlap=doc_chunk_overlap,
-            separator=" ",
-            paragraph_separator="\n\n",
-        )
-        self.code_parser = SentenceSplitter(
-            chunk_size=code_chunk_size,
-            chunk_overlap=code_chunk_overlap,
-            separator="\n",
-            paragraph_separator="\n\n",
-        )
-
-        all_texts = self.load_documents(
-            docs_paths, args.file_types, include_hidden=args.include_hidden
-        )
+        all_texts = self.load_documents(docs_paths, args.file_types)
         if not all_texts:
             print("No documents found")
             return
@@ -1349,8 +824,6 @@ Examples:
 
         if args.command == "list":
             self.list_indexes()
-        elif args.command == "remove":
-            self.remove_index(args.index_name, args.force)
         elif args.command == "build":
             await self.build_index(args)
         elif args.command == "search":
@@ -1362,15 +835,10 @@ Examples:
 
 
 def main():
-    import logging
-
     import dotenv
 
     dotenv.load_dotenv()
 
-    # Set clean logging for CLI usage
-    logging.getLogger().setLevel(logging.WARNING)  # Only show warnings and errors
-
     cli = LeannCLI()
     asyncio.run(cli.run())
 

packages/leann-core/src/leann/embedding_compute.py

@@ -57,8 +57,6 @@ def compute_embeddings(
         return compute_embeddings_mlx(texts, model_name)
     elif mode == "ollama":
         return compute_embeddings_ollama(texts, model_name, is_build=is_build)
-    elif mode == "gemini":
-        return compute_embeddings_gemini(texts, model_name, is_build=is_build)
     else:
         raise ValueError(f"Unsupported embedding mode: {mode}")
 
@@ -246,16 +244,6 @@ def compute_embeddings_openai(texts: list[str], model_name: str) -> np.ndarray:
     except ImportError as e:
         raise ImportError(f"OpenAI package not installed: {e}")
 
-    # Validate input list
-    if not texts:
-        raise ValueError("Cannot compute embeddings for empty text list")
-    # Extra validation: abort early if any item is empty/whitespace
-    invalid_count = sum(1 for t in texts if not isinstance(t, str) or not t.strip())
-    if invalid_count > 0:
-        raise ValueError(
-            f"Found {invalid_count} empty/invalid text(s) in input. Upstream should filter before calling OpenAI."
-        )
-
     api_key = os.getenv("OPENAI_API_KEY")
     if not api_key:
         raise RuntimeError("OPENAI_API_KEY environment variable not set")
@@ -275,16 +263,8 @@ def compute_embeddings_openai(texts: list[str], model_name: str) -> np.ndarray:
     print(f"len of texts: {len(texts)}")
 
     # OpenAI has limits on batch size and input length
-    max_batch_size = 800  # Conservative batch size because the token limit is 300K
+    max_batch_size = 1000  # Conservative batch size
     all_embeddings = []
-    # get the avg len of texts
-    avg_len = sum(len(text) for text in texts) / len(texts)
-    print(f"avg len of texts: {avg_len}")
-    # if avg len is less than 1000, use the max batch size
-    if avg_len > 300:
-        max_batch_size = 500
-
-    # if avg len is less than 1000, use the max batch size
 
     try:
         from tqdm import tqdm
@@ -670,83 +650,3 @@ def compute_embeddings_ollama(
     logger.info(f"Generated {len(embeddings)} embeddings, dimension: {embeddings.shape[1]}")
 
     return embeddings
-
-
-def compute_embeddings_gemini(
-    texts: list[str], model_name: str = "text-embedding-004", is_build: bool = False
-) -> np.ndarray:
-    """
-    Compute embeddings using Google Gemini API.
-
-    Args:
-        texts: List of texts to compute embeddings for
-        model_name: Gemini model name (default: "text-embedding-004")
-        is_build: Whether this is a build operation (shows progress bar)
-
-    Returns:
-        Embeddings array, shape: (len(texts), embedding_dim)
-    """
-    try:
-        import os
-
-        import google.genai as genai
-    except ImportError as e:
-        raise ImportError(f"Google GenAI package not installed: {e}")
-
-    api_key = os.getenv("GEMINI_API_KEY")
-    if not api_key:
-        raise RuntimeError("GEMINI_API_KEY environment variable not set")
-
-    # Cache Gemini client
-    cache_key = "gemini_client"
-    if cache_key in _model_cache:
-        client = _model_cache[cache_key]
-    else:
-        client = genai.Client(api_key=api_key)
-        _model_cache[cache_key] = client
-        logger.info("Gemini client cached")
-
-    logger.info(
-        f"Computing embeddings for {len(texts)} texts using Gemini API, model: '{model_name}'"
-    )
-
-    # Gemini supports batch embedding
-    max_batch_size = 100  # Conservative batch size for Gemini
-    all_embeddings = []
-
-    try:
-        from tqdm import tqdm
-
-        total_batches = (len(texts) + max_batch_size - 1) // max_batch_size
-        batch_range = range(0, len(texts), max_batch_size)
-        batch_iterator = tqdm(
-            batch_range, desc="Computing embeddings", unit="batch", total=total_batches
-        )
-    except ImportError:
-        # Fallback when tqdm is not available
-        batch_iterator = range(0, len(texts), max_batch_size)
-
-    for i in batch_iterator:
-        batch_texts = texts[i : i + max_batch_size]
-
-        try:
-            # Use the embed_content method from the new Google GenAI SDK
-            response = client.models.embed_content(
-                model=model_name,
-                contents=batch_texts,
-                config=genai.types.EmbedContentConfig(
-                    task_type="RETRIEVAL_DOCUMENT"  # For document embedding
-                ),
-            )
-
-            # Extract embeddings from response
-            for embedding_data in response.embeddings:
-                all_embeddings.append(embedding_data.values)
-        except Exception as e:
-            logger.error(f"Batch {i} failed: {e}")
-            raise
-
-    embeddings = np.array(all_embeddings, dtype=np.float32)
-    logger.info(f"Generated {len(embeddings)} embeddings, dimension: {embeddings.shape[1]}")
-
-    return embeddings

packages/leann-core/src/leann/embedding_server_manager.py

@@ -268,12 +268,8 @@ class EmbeddingServerManager:
             f"Terminating server process (PID: {self.server_process.pid}) for backend {self.backend_module_name}..."
         )
 
-        # Use simple termination first; if the server installed signal handlers,
-        # it will exit cleanly. Otherwise escalate to kill after a short wait.
-        try:
-            self.server_process.terminate()
-        except Exception:
-            pass
+        # Use simple termination - our improved server shutdown should handle this properly
+        self.server_process.terminate()
 
         try:
             self.server_process.wait(timeout=5)  # Give more time for graceful shutdown
@@ -282,10 +278,7 @@ class EmbeddingServerManager:
             logger.warning(
                 f"Server process {self.server_process.pid} did not terminate within 5 seconds, force killing..."
             )
-            try:
-                self.server_process.kill()
-            except Exception:
-                pass
+            self.server_process.kill()
             try:
                 self.server_process.wait(timeout=2)
                 logger.info(f"Server process {self.server_process.pid} killed successfully.")

packages/leann-core/src/leann/mcp.py

@@ -64,6 +64,19 @@ def handle_request(request):
                             "required": ["index_name", "query"],
                         },
                     },
+                    {
+                        "name": "leann_status",
+                        "description": "📊 Check the health and stats of your code indexes - like a medical checkup for your codebase knowledge!",
+                        "inputSchema": {
+                            "type": "object",
+                            "properties": {
+                                "index_name": {
+                                    "type": "string",
+                                    "description": "Optional: Name of specific index to check. If not provided, shows status of all indexes.",
+                                }
+                            },
+                        },
+                    },
                     {
                         "name": "leann_list",
                         "description": "📋 Show all your indexed codebases - your personal code library! Use this to see what's available for search.",
@@ -105,6 +118,15 @@ def handle_request(request):
                 ]
                 result = subprocess.run(cmd, capture_output=True, text=True)
 
+            elif tool_name == "leann_status":
+                if args.get("index_name"):
+                    # Check specific index status - for now, we'll use leann list and filter
+                    result = subprocess.run(["leann", "list"], capture_output=True, text=True)
+                    # We could enhance this to show more detailed status per index
+                else:
+                    # Show all indexes status
+                    result = subprocess.run(["leann", "list"], capture_output=True, text=True)
+
             elif tool_name == "leann_list":
                 result = subprocess.run(["leann", "list"], capture_output=True, text=True)
 

packages/leann-core/src/leann/registry.py

@@ -2,17 +2,11 @@
 
 import importlib
 import importlib.metadata
-import json
-import logging
-from pathlib import Path
-from typing import TYPE_CHECKING, Optional, Union
+from typing import TYPE_CHECKING
 
 if TYPE_CHECKING:
     from leann.interface import LeannBackendFactoryInterface
 
-# Set up logger for this module
-logger = logging.getLogger(__name__)
-
 BACKEND_REGISTRY: dict[str, "LeannBackendFactoryInterface"] = {}
 
 
@@ -20,7 +14,7 @@ def register_backend(name: str):
     """A decorator to register a new backend class."""
 
     def decorator(cls):
-        logger.debug(f"Registering backend '{name}'")
+        print(f"INFO: Registering backend '{name}'")
         BACKEND_REGISTRY[name] = cls
         return cls
 
@@ -45,54 +39,3 @@ def autodiscover_backends():
             # print(f"WARN: Could not import backend module '{backend_module_name}': {e}")
             pass
     # print("INFO: Backend auto-discovery finished.")
-
-
-def register_project_directory(project_dir: Optional[Union[str, Path]] = None):
-    """
-    Register a project directory in the global LEANN registry.
-
-    This allows `leann list` to discover indexes created by apps or other tools.
-
-    Args:
-        project_dir: Directory to register. If None, uses current working directory.
-    """
-    if project_dir is None:
-        project_dir = Path.cwd()
-    else:
-        project_dir = Path(project_dir)
-
-    # Only register directories that have some kind of LEANN content
-    # Either .leann/indexes/ (CLI format) or *.leann.meta.json files (apps format)
-    has_cli_indexes = (project_dir / ".leann" / "indexes").exists()
-    has_app_indexes = any(project_dir.rglob("*.leann.meta.json"))
-
-    if not (has_cli_indexes or has_app_indexes):
-        # Don't register if there are no LEANN indexes
-        return
-
-    global_registry = Path.home() / ".leann" / "projects.json"
-    global_registry.parent.mkdir(exist_ok=True)
-
-    project_str = str(project_dir.resolve())
-
-    # Load existing registry
-    projects = []
-    if global_registry.exists():
-        try:
-            with open(global_registry) as f:
-                projects = json.load(f)
-        except Exception:
-            logger.debug("Could not load existing project registry")
-            projects = []
-
-    # Add project if not already present
-    if project_str not in projects:
-        projects.append(project_str)
-
-        # Save updated registry
-        try:
-            with open(global_registry, "w") as f:
-                json.dump(projects, f, indent=2)
-            logger.debug(f"Registered project directory: {project_str}")
-        except Exception as e:
-            logger.warning(f"Could not save project registry: {e}")

packages/leann-mcp/README.md

@@ -4,29 +4,27 @@ Transform your development workflow with intelligent code assistance using LEANN
 
 ## Prerequisites
 
-Install LEANN globally for MCP integration (with default backend):
+**Step 1:** First, complete the basic LEANN installation following the [📦 Installation guide](../../README.md#installation) in the root README:
 
 ```bash
-uv tool install leann-core --with leann
+uv venv
+source .venv/bin/activate
+uv pip install leann
 ```
-This installs the `leann` CLI into an isolated tool environment and includes both backends so `leann build` works out-of-the-box.
+
+**Step 2:** Install LEANN globally for MCP integration:
+```bash
+uv tool install leann-core
+```
+
+This makes the `leann` command available system-wide, which `leann_mcp` requires.
 
 ## 🚀 Quick Setup
 
-Add the LEANN MCP server to Claude Code. Choose the scope based on how widely you want it available. Below is the command to install it globally; if you prefer a local install, skip this step:
+Add the LEANN MCP server to Claude Code:
 
 ```bash
-# Global (recommended): available in all projects for your user
-claude mcp add --scope user leann-server -- leann_mcp
-```
-
-- `leann-server`: the display name of the MCP server in Claude Code (you can change it).
-- `leann_mcp`: the Python entry point installed with LEANN that starts the MCP server.
-
-Verify it is registered globally:
-
-```bash
-claude mcp list | cat
+claude mcp add leann-server -- leann_mcp
 ```
 
 ## 🛠️ Available Tools
@@ -35,36 +33,27 @@ Once connected, you'll have access to these powerful semantic search tools in Cl
 
 - **`leann_list`** - List all available indexes across your projects
 - **`leann_search`** - Perform semantic searches across code and documents
-
+- **`leann_ask`** - Ask natural language questions and get AI-powered answers from your codebase
 
 ## 🎯 Quick Start Example
 
 ```bash
-# Add locally if you did not add it globally (current folder only; default if --scope is omitted)
-claude mcp add leann-server -- leann_mcp
-
 # Build an index for your project (change to your actual path)
-# See the advanced examples below for more ways to configure indexing
-# Set the index name (replace 'my-project' with your own)
-leann build my-project --docs $(git ls-files)
+leann build my-project --docs ./
 
 # Start Claude Code
 claude
 ```
 
-## 🚀 Advanced Usage Examples to build the index
+## 🚀 Advanced Usage Examples
 
 ### Index Entire Git Repository
 ```bash
-# Index all tracked files in your Git repository.
-# Note: submodules are currently skipped; we can add them back if needed.
+# Index all tracked files in your git repository, note right now we will skip submodules, but we can add it back easily if you want
 leann build my-repo --docs $(git ls-files) --embedding-mode sentence-transformers --embedding-model all-MiniLM-L6-v2 --backend hnsw
 
-# Index only tracked Python files from Git.
+# Index only specific file types from git
 leann build my-python-code --docs $(git ls-files "*.py") --embedding-mode sentence-transformers --embedding-model all-MiniLM-L6-v2 --backend hnsw
-
-# If you encounter empty requests caused by empty files (e.g., __init__.py), exclude zero-byte files. Thanks @ww2283 for pointing [that](https://github.com/yichuan-w/LEANN/issues/48) out
-leann build leann-prospec-lig --docs $(find ./src -name "*.py" -not -empty) --embedding-mode openai --embedding-model text-embedding-3-small
 ```
 
 ### Multiple Directories and Files
@@ -92,7 +81,7 @@ leann build docs-and-configs --docs $(git ls-files "*.md" "*.yml" "*.yaml" "*.js
 ```
 
 
-## **Try this in Claude Code:**
+**Try this in Claude Code:**
 ```
 Help me understand this codebase. List available indexes and search for authentication patterns.
 ```
@@ -101,7 +90,6 @@ Help me understand this codebase. List available indexes and search for authenti
   <img src="../../assets/claude_code_leann.png" alt="LEANN in Claude Code" width="80%">
 </p>
 
-If you see a prompt asking whether to proceed with LEANN, you can now use it in your chat!
 
 ## 🧠 How It Works
 
@@ -137,11 +125,3 @@ To remove LEANN
 ```
 uv pip uninstall leann leann-backend-hnsw leann-core
 ```
-
-To globally remove LEANN (for version update)
-```
-uv tool list | cat
-uv tool uninstall leann-core
-command -v leann || echo "leann gone"
-command -v leann_mcp || echo "leann_mcp gone"
-```

packages/leann/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "leann"
-version = "0.3.1"
+version = "0.2.8"
 description = "LEANN - The smallest vector index in the world. RAG Everything with LEANN!"
 readme = "README.md"
 requires-python = ">=3.9"

packages/wechat-exporter/__init__.py

@@ -1 +0,0 @@
-__all__ = []

packages/wechat-exporter/main.py

@@ -136,9 +136,5 @@ def export_sqlite(
     connection.commit()
 
 
-def main():
-    app()
-
-
 if __name__ == "__main__":
-    main()
+    app()

pyproject.toml

@@ -10,10 +10,11 @@ requires-python = ">=3.9"
 dependencies = [
     "leann-core",
     "leann-backend-hnsw",
-    "typer>=0.12.3",
     "numpy>=1.26.0",
     "torch",
     "tqdm",
+    "flask",
+    "flask_compress",
     "datasets>=2.15.0",
     "evaluate",
     "colorama",
@@ -64,7 +65,9 @@ test = [
     "pytest>=7.0",
     "pytest-timeout>=2.0",
     "llama-index-core>=0.12.0",
+    "llama-index-readers-file>=0.4.0",
     "python-dotenv>=1.0.0",
+    "sentence-transformers>=2.2.0",
 ]
 
 diskann = [
@@ -81,11 +84,6 @@ documents = [
 
 [tool.setuptools]
 py-modules = []
-packages = ["wechat_exporter"]
-package-dir = { "wechat_exporter" = "packages/wechat-exporter" }
-
-[project.scripts]
-wechat-exporter = "wechat_exporter.main:main"
 
 
 [tool.uv.sources]
@@ -96,8 +94,13 @@ leann-backend-hnsw = { path = "packages/leann-backend-hnsw", editable = true }
 [tool.ruff]
 target-version = "py39"
 line-length = 100
-extend-exclude = ["third_party"]
-
+extend-exclude = [
+    "third_party",
+    "*.egg-info",
+    "__pycache__",
+    ".git",
+    ".venv",
+]
 
 [tool.ruff.lint]
 select = [
@@ -120,12 +123,21 @@ ignore = [
     "RUF012", # mutable class attributes should be annotated with typing.ClassVar
 ]
 
+[tool.ruff.lint.per-file-ignores]
+"test/**/*.py" = ["E402"]      # module level import not at top of file (common in tests)
+"examples/**/*.py" = ["E402"]  # module level import not at top of file (common in examples)
+
 [tool.ruff.format]
 quote-style = "double"
 indent-style = "space"
 skip-magic-trailing-comma = false
 line-ending = "auto"
 
+[dependency-groups]
+dev = [
+    "ruff>=0.12.4",
+]
+
 [tool.lychee]
 accept = ["200", "403", "429", "503"]
 timeout = 20

sky/leann-build.yaml

@@ -1,76 +0,0 @@
-name: leann-build
-
-resources:
-  # Choose a GPU for fast embeddings (examples: L4, A10G, A100). CPU also works but is slower.
-  accelerators: L4:1
-  # Optionally pin a cloud, otherwise SkyPilot will auto-select
-  # cloud: aws
-  disk_size: 100
-
-envs:
-  # Build parameters (override with: sky launch -c leann-gpu sky/leann-build.yaml -e key=value)
-  index_name: my-index
-  docs: ./data
-  backend: hnsw               # hnsw | diskann
-  complexity: 64
-  graph_degree: 32
-  num_threads: 8
-  # Embedding selection
-  embedding_mode: sentence-transformers   # sentence-transformers | openai | mlx | ollama
-  embedding_model: facebook/contriever
-  # Storage/latency knobs
-  recompute: true             # true => selective recomputation (recommended)
-  compact: true               # for HNSW only
-  # Optional pass-through
-  extra_args: ""
-  # Rebuild control
-  force: true
-
-# Sync local paths to the remote VM. Adjust as needed.
-file_mounts:
-  # Example: mount your local data directory used for building
-  ~/leann-data: ${docs}
-
-setup: |
-  set -e
-  # Install uv (package manager)
-  curl -LsSf https://astral.sh/uv/install.sh | sh
-  export PATH="$HOME/.local/bin:$PATH"
-
-  # Ensure modern libstdc++ for FAISS (GLIBCXX >= 3.4.30)
-  sudo apt-get update -y
-  sudo apt-get install -y libstdc++6 libgomp1
-  # Also upgrade conda's libstdc++ in base env (Skypilot images include conda)
-  if command -v conda >/dev/null 2>&1; then
-    conda install -y -n base -c conda-forge libstdcxx-ng
-  fi
-
-  # Install LEANN CLI and backends into the user environment
-  uv pip install --upgrade pip
-  uv pip install leann-core leann-backend-hnsw leann-backend-diskann
-
-run: |
-  export PATH="$HOME/.local/bin:$PATH"
-  # Derive flags from env
-  recompute_flag=""
-  if [ "${recompute}" = "false" ] || [ "${recompute}" = "0" ]; then
-    recompute_flag="--no-recompute"
-  fi
-  force_flag=""
-  if [ "${force}" = "true" ] || [ "${force}" = "1" ]; then
-    force_flag="--force"
-  fi
-
-  # Build command
-  python -m leann.cli build ${index_name} \
-    --docs ~/leann-data \
-    --backend ${backend} \
-    --complexity ${complexity} \
-    --graph-degree ${graph_degree} \
-    --num-threads ${num_threads} \
-    --embedding-mode ${embedding_mode} \
-    --embedding-model ${embedding_model} \
-    ${recompute_flag} ${force_flag} ${extra_args}
-
-  # Print where the index is stored for downstream rsync
-  echo "INDEX_OUT_DIR=~/.leann/indexes/${index_name}"