Agent memory you can 'git blame' and how OKF4net grew into a zero-dependency .NET toolkit
Most answers to "give my AI agent long-term memory" hand you an opaque vector store: embeddings in a database you can't read, can't diff, and can't easily redact. I wanted the opposite : memory I can open in an editor, review in a pull request, and 'git blame' line by line. That itch is what pushed OKF4net from a single library into a small toolkit over the last few months. It just reached v0.5 ,…
OKF4net is an independent, zero-dependency .NET implementation of Google's Open Knowledge Format (OKF), which represents knowledge as a directory of markdown files with YAML frontmatter, cross-linked like a wiki and versioned in git. The project has evolved from a single library into a small toolkit, reaching v0.5. The main feature of OKF4net is memory capture, which allows agents to remember facts in a plain markdown format that can be opened in an editor, reviewed in a pull request, and git blamed line by line.
This memory capture is opt-in and deterministic, with no extra LLM call, ensuring determinism and easy review. The captured memory is just files in the bundle, making it possible to track changes, diff across commits, and review the whole thing in a PR. The toolkit also includes OkfBundleTools for reading, browsing, graphing, searching, writing, appending, validating, and more; OkfContextProvider for injecting bundle context into each turn of the Microsoft Agent Framework; and an MCP server for exposing the bundle to other editors.
The project is open source under LGPL-3.0-or-later, welcoming first contributions with no prior OKF knowledge required.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
