liliu-z / stashbase

What is stashbase?

StashBase is a local knowledge base that keeps your notes, papers, and research continuously indexed and searchable for AI workflows. It runs as a desktop app on macOS (with Windows/Linux in development) and exposes your library to Claude Code, ChatGPT, and other MCP-aware tools through semantic and keyword search. Notes are stored as HTML or Markdown, and the system automatically re-indexes when files change.

Why is it gaining traction?

The appeal lies in how it bridges personal knowledge management with AI tooling—your notes become context for coding agents without manual uploads or copy-pasting. The MCP integration means you can query your library directly from Claude or ChatGPT using natural language, and the in-app terminal lets Claude Code work inside your knowledge space with immediate indexing. The PDF-to-HTML pipeline is particularly useful for researchers handling academic papers.

Who should use this?

Researchers and knowledge workers drowning in PDFs and notes would benefit most. Developers using Claude Code or Codex who want their accumulated notes to inform AI assistance will find this valuable. Anyone maintaining a personal knowledge base across Markdown or HTML files and wanting AI-powered retrieval without cloud dependencies could use it.

Verdict

StashBase shows promise as a local-first knowledge layer for AI workflows, but it's early-stage software with only 14 stars and a 0.95% credibility score. The feature set is solid for its intended use case, though the sparse community and documentation warrant caution for production deployments. If you're comfortable with alpha software and want to experiment with AI-augmented note retrieval on your own machine, it's worth exploring; otherwise, wait for more maturity.