
Quorum AI
io.github.quorum-dogPublished by GitHub account quorum-dogv1.0.0Updated Oct 25, 2026
Compare AI perspectives with configurable Modes, estimate deliberations, and retrieve run receipts.
Overview
Quorum AI is a remote MCP server that lets an assistant run multi-model deliberations, compare AI perspectives, estimate costs, and fetch run receipts.
- What you can do
- List the Quorum Modes available to the connected account and request a cost estimate before running anything. Run a selected Mode so several models critique, judge and synthesize an answer, surfacing agreement and disagreement. Retrieve a receipt for a run with participating models, cost, scores and latency where available. Create, validate, test, certify and publish Modes, subject to account permissions.
- When to use it
- Useful when you want a second opinion on architecture choices, implementation plans, assumptions or build-versus-buy decisions, and want several models compared rather than one answer. Best for users who already have a Quorum account and want the deliberation workflow inside their assistant.
- Requirements
- A remote endpoint reached over streamable HTTP; no local package or pasted API key. A Quorum account is required, and OAuth sign-in must be completed in the client. Client support for custom remote MCP connectors varies by client and plan.
Installation
In SourceWeft
- Open Quorum AI in the dashboard and add it to a workspace.
- Connect the required credentials. They are stored privately and never shown on public pages.
- Enable the server for the chats that should use its tools.
Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.
Use with an agent
Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.
Read https://sourceweft.com/mcp/io.github.quorum-dog%2Fquorum/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.Other MCP clients
Add this to your client's mcpServers config. The client completes OAuth sign-in on first connect.
{
"mcpServers": {
"quorum-ai": {
"type": "http",
"url": "https://www.quorum.dog/mcp"
}
}
}README
Quorum MCP
[Quorum AI]AI’s second opinion.
Quorum uses configurable Modes to gather and synthesize responses from AI models. Depending on the selected Mode, a deliberation can include multiple models, critique, judging, and synthesis. Results can help users examine different perspectives, including areas of agreement and disagreement.
Use Quorum inside your AI assistant to explore assumptions, compare architecture choices, evaluate implementation plans, and consider build-versus-buy decisions. Outputs are AI-generated and may be incomplete or incorrect.
What the connector offers
- List the Quorum Modes available to your account.
- Request an estimate before running a deliberation.
- Run a selected Mode to compare AI perspectives.
- Retrieve a receipt with participating models, cost, scores, and latency where available.
- Create, validate, test, certify, and publish Modes, subject to your account permissions.
Connect with Cursor
A Quorum account is required. Install the plugin and complete Quorum’s OAuth sign-in when Cursor prompts you. The plugin connects to https://www.quorum.dog/mcp using remote HTTP. No local server, package installation, or pasted API key is needed for this OAuth setup.
For a direct connection before marketplace approval, add this configuration to your Cursor MCP settings (.cursor/mcp.json for a project, or your global Cursor MCP configuration):
Enable Quorum in Cursor’s MCP settings and complete sign-in. Ask Cursor to list your Modes to check the connection.
Connect with VS Code
Add the following to .vscode/mcp.json in your project. Enable the server and complete Quorum’s OAuth sign-in when prompted.
You can also use the MCP: Add Server command and select HTTP, then enter the endpoint above.
Connect with Claude or ChatGPT
Where your client supports custom remote MCP connectors, use https://www.quorum.dog/mcp and complete OAuth sign-in. Availability and setup depend on the client and account plan. See the Quorum MCP guide for client-specific instructions.
Try it
- “List the Quorum Modes available to me. Don’t run a deliberation.”
- “Estimate a Quorum review of whether our five-person consultancy should build or buy its client onboarding system. Don’t run the panel yet.”
- After reviewing the estimate: “Run that Quorum deliberation using the selected Mode.”
- “Pull the receipt for that run. Show participating models, cost, judge scores, and latency where available.”
Other useful questions:
- “Get a Quorum estimate for comparing a modular monolith with microservices for this project. Don’t run yet.”
- “Use Quorum to review this implementation plan after I approve the estimate. Identify assumptions and tradeoffs.”
Estimates can return unavailable pricing fields; an unavailable price is not a zero-cost quote. Deliberations and other execution tools may incur usage charges. Follow the tools’ approval requirements and the user’s spending instructions. Report missing receipt fields as unavailable.
Data and permissions
Prompts and context sent to Quorum are processed by the Quorum service and the model providers used for the selected Mode. Send only the context needed for the requested review. The connector also exposes tools that can save or publish Modes; use those only when requested and permitted by the connected account.
This repository contains public connector configuration, setup documentation, and brand assets. Quorum’s hosted application runs separately.
Source: README.md at commit f12dafa
Tools
0Version history
1- v1.0.0LatestOct 18, 2026
Score
75% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 daysRequired25 pts
- Auth declaredRequired10 pts
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts
Declare a license in the repository or the registry entry.