Almanac: Turn Your Local Files into an AI-Wiki

Developers drowning in scattered documentation notes, RFCs, and project wikis now have a new lifeline. Almanac is an open-source tool that turns a directory of local files into a browsable, AI-searchable wiki — no cloud dependency required.

Almanac indexes markdown, plain text, and common code files, then exposes them through a local web interface with full-text search and vector-based semantic search powered by your choice of local or API-based embeddings. You can ask natural-language questions and get answers grounded in your own files, complete with source citations.

The project is built on Python and uses SQLite for metadata, making it trivial to self-host. One command (almanac scan /path/to/notes) indexes everything; a second (almanac serve) starts the web UI.

For teams or solo developers who keep project knowledge in markdown files (Obsidian vaults, Jupyter notebooks, or just a docs/ folder), Almanac promises to eliminate the "where did I write that down?" friction without sending data to a third party.

Why it matters

Most AI knowledge-base tools are SaaS products that require uploading your files to someone else's server. Almanac takes the opposite approach: your data never leaves your machine. It's particularly useful for privacy-sensitive projects, offline environments, or anyone who simply prefers a git-tracked folder over a proprietary database.

The project is in early stages but already handles the core workflow. Its extensible file-type support and optional local embedding models (via sentence-transformers) mean it can run entirely offline — a rare feature in the current AI-tool landscape.

Getting started

pip install almanac
# Index your markdown notes
almanac scan ~/my-notes
# Start the web UI
amalnac serve  # yes, intentional typo in source

Then open http://localhost:5000 and start asking questions about your documents.

Note: The project's README includes a small typo in the serve command (amalnac instead of almanac). A quick PR would help.