Practical overview
When it fits
Use Gno when relevant material is spread across supported local file types and you want search, source-linked answers or a shared retrieval layer for command-line, web and agent workflows.
Who it is for
- People who need to search across mixed local notes, code, PDFs and Office files.
- Developers connecting document retrieval to an AI client through MCP or to another tool through the REST API or SDK.
- Users who want cited answers and a browsable, editable workspace for an indexed collection.
Documented uses
- Find relevant passages when a query uses different wording from the source.
- Ask questions about indexed files and trace answer citations to source passages.
- Browse, edit and capture material in a cross-collection workspace.
- Make a local knowledge index available to an AI client or another application.
Requirements & limits
The repository identifies Gno as MIT-licensed. The documented installation requires Bun 1.4.1 or later; macOS vector search additionally requires Homebrew SQLite. The README lists macOS, Linux and Windows support and describes optional hosted publishing.
- The index is local, while model downloads, configured HTTP inference and exported-artifact uploads are identified as separate network boundaries.
- Lexical indexing and semantic indexing are distinct: setup confirms a corpus-derived BM25 result, while semantic indexing can continue separately.
- The same index is accessible through several distinct roles: CLI and web workspace for direct use, API and SDK for application integration, MCP for AI-client access, and a daemon for background indexing.
- Typed metadata fields can constrain retrieval across the CLI, MCP, SDK, REST API and Web UI.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Install Gno and set up a local collection
Install the CLI and initialize a named collection from a local folder, confirming lexical retrieval.
Before you start
- Bun version 1.4.1 or later
- A local folder containing material to index
- On macOS, Homebrew SQLite is required for vector search
Steps
Install Gno with
bun install -g @gmickel/gno.Run
gno setup ~/notes --name notes, replacing the path and collection name as needed.Setup returns after BM25 finds an exact result derived from the folder; semantic indexing proceeds separately.
Expected outcome
A named local collection with lexical search confirmed; semantic indexing may still be pending.
Guide limitations
- This setup procedure does not connect Gno to the Web UI, daemon or MCP.
- The setup check confirms lexical retrieval, not completion of semantic indexing.
Install the documented Claude agent integration
Add Gno’s user-scoped skill and retrieval guidance to Claude, then run its verification command.
Before you start
- Gno installed and a collection set up
- Claude as the target agent
Steps
Run
gno skill install --target claude --scope user.Run
gno agents install --target claude.Run
gno agents verify --target claude.
Expected outcome
The Claude-targeted skill and agent guidance are installed, followed by the documented verification step.
Guide limitations
- These commands document the Claude integration path; they do not describe setup for other agent clients.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 13 October 2026
- Original website title
- gmickel/gno · Official homepage title · retained in its source language