Roux MCP server
Give your AI the human context it's missing.
Connect Roux to the AI where work already happens. Launch structured interviews, find existing Roux, track responses, and carry synthesized findings back into the decision.
Installation
Choose your AI client
The guided pages give the AI everything it needs to walk you through setup. You can also configure any remote MCP client manually.
Remote MCP endpoint
https://www.helloroux.com/mcp- 01
Open connector settings
Open the custom MCP or connector settings in your AI client.
- 02
Add the remote server
Name it Roux and paste the Streamable HTTP endpoint above. Do not add authentication headers.
- 03
Authorize one workspace
Connect through OAuth, sign in to Roux, and choose the workspace this client may access.
- 04
Run a read-only test
Start a new conversation and ask the client to list recent Roux or search for one by title.
MCP tools
From missing context to a full synthesis
Find the right Roux
List recent Roux, search titles and goals, and continue through large workspaces with cursor pagination.
list_rouxsearch_rouxCollect human context
Create a structured Roux, share its response link, and track completion without exposing respondent identities.
create_rouxget_roux_statusTurn responses into decisions
Generate a synthesis, read the findings in conversation, and open the full report in the Roux dashboard.
generate_roux_synthesisget_roux_synthesisConnect the rest of your stack
Bring GitHub, Linear, and Slack context into Roux, then let your AI carry the synthesized result back to where work happens.