The answer is hidden somewhere across your sources. passageway finds it.

passageway searches all your sources at once, wiki, notes and documentation, and gives your agent exactly the passages that answer the question, each linked to its source.

Install and run Write a connector

• On premise. Your knowledge stays on your infrastructure.

  Run it yourself

• Connected. Confluence, Trilium, Markdown and Wikipedia work out of the box.

  See the connectors

• Open. A documented protocol lets you add any source or embedding model, in any language.

  Read the protocol

Free for private use. Companies need a licence. Details

What it does

passageway is the retrieval half of a RAG system, served as an MCP tool. It does no generation: the agent that calls it reads the passages and writes the answer. A relevance gate returns nothing instead of padding a prompt with weak matches. On the sample corpus the right article is among the top three for 88.6% (95% interval 79–94%) of 70 questions: the method and its limits. How it works

Start

• Install in Claude Code: the plugin sets up the server, the sources and the hook for you.
• Add a source: Markdown, Trilium, Confluence, Wikipedia, llms.txt, Sitemap.
• Measure your own notes: passageway eval on your questions, and choose a model.

What passageway stands for and the roadmap, for what is planned.