Metadata-Version: 2.4
Name: humona
Version: 0.0.1
Summary: Human-inspired memory for AI agents — early name registration; install from GitHub until the packaged release lands
Author: Mesut Can Demir
License: MIT
Project-URL: Homepage, https://github.com/mesutcandemir39/humona
Project-URL: Source, https://github.com/mesutcandemir39/humona
Project-URL: Changelog, https://github.com/mesutcandemir39/humona/blob/main/CHANGELOG.md
Keywords: humona,memory,claude-code,agent,mcp,obsidian
Classifier: Development Status :: 2 - Pre-Alpha
Classifier: Environment :: Console
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Software Development :: Libraries
Requires-Python: >=3.8
Description-Content-Type: text/markdown
License-File: LICENSE
Dynamic: license-file

# humona

**Human-inspired memory for AI agents.** Markdown notes in a git repository you own — no
database, no service, no inference to remember.

This PyPI release is **0.0.1: the name, registered ahead of the packaged release.** The engine
itself is complete, tested on macOS, Linux and Windows, and in daily use — but it is installed
from its repository, because a pip install cannot yet carry the hooks, skills and installer
that make it work.

## Install the working engine

As a Claude Code plugin:

```
/plugin marketplace add mesutcandemir39/humona
/plugin install humona@humona
```

Standalone:

```bash
git clone https://github.com/mesutcandemir39/humona.git ~/humona-src
python3 ~/humona-src/setup/install.py
```

Everything — the measurements, the architecture, the SWOT and the comparison with other memory
systems — is at **https://github.com/mesutcandemir39/humona**.

MIT licensed.
