### Development Installation and Testing

Source: https://github.com/iwaan10000vr/vecmemori/blob/main/README.md

Install the package in development mode with optional dependencies and run tests. Build source and wheel distributions, and check them using twine.

```bash
pip install -e ".[dev,ja]"
python -m pytest -q
python -m build --sdist --wheel
python -m twine check dist/*
check-manifest
```

--------------------------------

### Set Up Development Environment

Source: https://github.com/iwaan10000vr/vecmemori/blob/main/CONTRIBUTING.md

Clone the repository, create and activate a virtual environment, install development dependencies, and download necessary models.

```bash
git clone https://github.com/iwaan10000vr/vecmemori
cd vecmemori
python3 -m venv .venv
source .venv/bin/activate
pip install -e ".[dev,ja]"
bash scripts/download_model.sh
```

--------------------------------

### Verify Japanese FTS5 Tokenizer Installation

Source: https://github.com/iwaan10000vr/vecmemori/blob/main/README.md

After installing the Japanese extra, run this Python command to verify that the Japanese FTS5 tokenizer is correctly installed and available. This confirms setup for Japanese text processing.

```python
from vecmemori._tokenizer import has_tokenizer; print('Japanese FTS5:', has_tokenizer())
```

--------------------------------

### Install vecmemori and extras

Source: https://github.com/iwaan10000vr/vecmemori/blob/main/README.md

Install the vecmemori package with optional extras for Japanese FTS5 tokenization, Hermes Agent integration, or both. Use this to set up the library for specific functionalities.

```bash
pip install vecmemori
pip install vecmemori[ja]       # Japanese FTS5 tokenization with fugashi + unidic-lite
pip install vecmemori[hermes]   # adapter dependencies for Hermes Agent environments
pip install vecmemori[all]      # Japanese + Hermes adapter dependencies
```

--------------------------------

### Quick Start: Add and Retrieve Facts

Source: https://github.com/iwaan10000vr/vecmemori/blob/main/README.md

Initialize MemoryStore, add facts with categories, and then use FactRetriever to search for relevant information. This demonstrates the basic workflow for storing and retrieving facts.

```python
from vecmemori import MemoryStore, FactRetriever

store = MemoryStore(db_path="memory.db")
store.add_fact("User prefers dark mode in all applications", category="user_pref")
store.add_fact("Project uses pytest for testing", category="project")

retriever = FactRetriever(store=store)
results = retriever.search("testing preferences", limit=5)

for r in results:
    print(f"[{r['trust_score']:.2f}] {r['content']}")
```

--------------------------------

### Install Vecmemori from GitHub

Source: https://github.com/iwaan10000vr/vecmemori/blob/main/README.md

Clone the repository, install the Python package with Hermes and Japanese language support, download the model, and link it as a Hermes plugin. Configure Hermes to use vecmemori as its memory provider and set the embedding model path.

```bash
git clone https://github.com/iwaan10000vr/vecmemori.git
cd vecmemori
python -m pip install -e "[hermes,ja]"
bash scripts/download_model.sh

mkdir -p ~/.hermes/plugins
ln -s "$(pwd)/src/vecmemori/hermes" ~/.hermes/plugins/vecmemori
hermes config set memory.provider vecmemori
hermes config set plugins.vecmemori.embedding_model "$HOME/.cache/vecmemori/models/ruri-v3-310m"
```

--------------------------------

### Run Tests and Build Distribution

Source: https://github.com/iwaan10000vr/vecmemori/blob/main/CONTRIBUTING.md

Execute tests using pytest, build source distributions and wheels, and check the distribution package.

```bash
python -m pytest -q
python -m build --sdist --wheel
python -m twine check dist/*
check-manifest
```

--------------------------------

### Optional Local-First Settings for Vecmemori

Source: https://github.com/iwaan10000vr/vecmemori/blob/main/README.md

Configure vecmemori for local-first operation, including automatic fact extraction, specifying the fact storage language, disabling the retrieval planner, and controlling trust in remote embedding model code.

```bash
hermes config set plugins.vecmemori.auto_extract true
hermes config set plugins.vecmemori.fact_storage_language ja   # language used to write stored fact content
hermes config set plugins.vecmemori.retrieval_planner false
hermes config set plugins.vecmemori.embedding_trust_remote_code false
```

--------------------------------

### Configure Embedding Model and Rebuild Embeddings

Source: https://github.com/iwaan10000vr/vecmemori/blob/main/README.md

Set the configuration for embedding dimensions and prefixes, specify the local model path, and then initialize a MemoryStore and rebuild all embeddings. This is useful when swapping to a different embedding model.

```python
from vecmemori._embedder import set_config, set_model_path

set_config(dimension=384, query_prefix="", doc_prefix="")
set_model_path("/path/to/all-MiniLM-L6-v2")

from vecmemori import MemoryStore
store = MemoryStore("memory.db")
store.rebuild_all_embeddings()
```

--------------------------------

### Download default embedding model

Source: https://github.com/iwaan10000vr/vecmemori/blob/main/README.md

Execute this bash script to download the default SentenceTransformer-compatible embedding model. This is a prerequisite for using vecmemori's semantic search capabilities.

```bash
bash scripts/download_model.sh
```

--------------------------------

### MIT License for vecmemori

Source: https://github.com/iwaan10000vr/vecmemori/blob/main/THIRD_PARTY_NOTICES.md

This is the MIT License text for the vecmemori project. It grants broad permissions for use, modification, and distribution.

```text
MIT License

Copyright (c) 2026 iwaan

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
```

--------------------------------

### Python API: Add Fact, Record Feedback, and Rebuild Embeddings

Source: https://github.com/iwaan10000vr/vecmemori/blob/main/README.md

Demonstrates adding a fact with tags, recording user feedback to adjust trust scores, and rebuilding all embeddings. This is useful for managing fact reliability and updating the embedding index.

```python
from vecmemori import MemoryStore, FactRetriever

store = MemoryStore(db_path="memory.db")
fact_id = store.add_fact("GPU: RTX 5060 Ti 16GB", category="tool", tags="hardware,gpu")
store.record_feedback(fact_id=fact_id, helpful=True)

retriever = FactRetriever(store=store)
results = retriever.search("gpu memory", limit=5)

store.rebuild_all_embeddings()  # after changing embedding model/config
store.close()
```

--------------------------------

### BSD-3-Clause License for NumPy

Source: https://github.com/iwaan10000vr/vecmemori/blob/main/THIRD_PARTY_NOTICES.md

This is the BSD-3-Clause License text for the NumPy library. It outlines the terms for redistribution and use of NumPy.

```text
Copyright (c) 2005-2024, NumPy Developers.
All rights reserved.

Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:

1. Redistributions of source code must retain the above copyright notice,
   this list of conditions and the following disclaimer.

2. Redistributions in binary form must reproduce the above copyright notice,
   this list of conditions and the following disclaimer in the documentation
   and/or other materials provided with the distribution.

3. Neither the name of the copyright holder nor the names of its
   contributors may be used to endorse or promote products derived from
   this software without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.