============
Installation
============

System Requirements
-------------------

- **OS**: Linux (Ubuntu 22.04+ recommended). WSL2 on Windows works too.
- **Python**: 3.10 or later
- **RAM**: 8 GB minimum, 16 GB recommended for AI analysis
- **GPU**: Optional - NVIDIA with 6+ GB VRAM accelerates LLM inference

Step 1: Install MemGuard
------------------------

.. code-block:: bash

   git clone https://github.com/memory-analyzer/memguard.git
   cd memguard/memguard_release
   python3 -m venv .venv
   source .venv/bin/activate
   pip install -e .

Step 2: Install Detection Tools
-------------------------------

.. code-block:: bash

   # Core tools
   sudo apt install -y valgrind gcc g++ gdb cppcheck clang-tidy heaptrack

   # Facebook Infer
   VERSION=v1.2.0
   curl -sSL "https://github.com/facebook/infer/releases/download/$VERSION/infer-linux-x86_64-$VERSION.tar.xz" \
     | sudo tar -C /opt -xJ
   sudo ln -sf /opt/infer-linux-x86_64-$VERSION/bin/infer /usr/local/bin/infer

   # Mozilla rr (build from source for latest CPU support)
   sudo apt install -y cmake ninja-build pkg-config capnproto libcapnp-dev
   cd /tmp && git clone https://github.com/rr-debugger/rr.git
   cd rr && mkdir build && cd build
   cmake -G Ninja .. -DCMAKE_BUILD_TYPE=Release -Ddisable32bit=ON
   ninja -j$(nproc) && sudo ninja install
   sudo sysctl kernel.perf_event_paranoid=1

   # Neuro-symbolic dependencies
   pip install z3-solver tree-sitter tree-sitter-c

Step 3: Install AI Backend
--------------------------

.. code-block:: bash

   # Ollama - runs locally, no cloud
   curl -fsSL https://ollama.com/install.sh | sh
   ollama pull qwen2.5-coder:14b-instruct-q4_K_M

.. note::

   All AI inference runs locally on your machine. No data leaves your network.
   The 14B model needs ~9 GB disk and 8 GB VRAM.

Step 4: Verify
--------------

.. code-block:: bash

   memguard doctor