Remyx from the command line

Recommendations, search, and Outrider in your scripts and CI. The same engine that powers Studio, callable from your terminal.

$ pip install remyxai

// 01 common_commands

A few things you can run

interests from-repo

$ remyxai interests from-repo \
    github.com/org/repo \
    --automate review

Profile a GitHub repo into a Research Interest, and provision Outrider on it in the same step.

papers digest

$ remyxai papers digest \
    --period today

Today's ranked recommendations, grouped by interest, across papers, repos, and models.

search query

$ remyxai search query \
    "speculative decoding" \
    --docker

Semantic search across the Remyx index, filtered to resources with a ready Docker environment.

outrider init

$ remyxai outrider init \
    --repo owner/name \
    --auto-interest

Install Outrider on a repository and auto-create its Research Interest in one command.

# authenticate first with export REMYXAI_API_KEY=...

// 02 other_ways_in

studio

The visual interface for experiment outcomes, discovery, and projects. Open Studio →

mcp server

An AI-native interface for Claude Code and other MCP clients, at mcp.remyx.ai. MCP docs →

rest api

Programmatic access to experiments, recommendations, and projects. API reference →

github action

Outrider runs as a Continuous AI action on your repositories. About Outrider →

Install the CLI and get your first recommendation in minutes.