Agience Core

v1.0.0Knowledge & Memorystable

The operating system that AI workflows trust. Structure messy inputs. Establish identity. Track provenance. Build the durable data layer your AI workflows need to be trusted, reusable, and auditable.

ai-agentsfastapiknowledge-managementmcpmcp-gateway
Share:
39
Stars
0
Downloads
0
Weekly
0/5

What is Agience Core?

Agience Core is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to operating system that ai workflows trust. structure messy inputs. establish identity. track provenance. build the durable data layer your ai workflows need to be trusted, reusable, and auditable.

The operating system that AI workflows trust. Structure messy inputs. Establish identity. Track provenance. Build the durable data layer your AI workflows need to be trusted, reusable, and auditable.

This server falls under the Knowledge & Memory category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • The operating system that AI workflows trust. Structure mess

Use Cases

Structure unstructured AI workflow inputs and establish identity tracking. Build auditable, reusable data layers with full provenance tracking.
Agience

Maintainer

LicenseNOASSERTION
Languagepython
Versionv1.0.0
UpdatedMay 8, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx agience-core

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Agience Core

Agience Core is a self-hostable platform that acts as an operating system for AI workflows, turning raw LLM output into durable, governed knowledge artifacts with identity, provenance, and version history. It ships as an MCP server exposing tools to VS Code, Claude Desktop, and Cursor, while simultaneously consuming external MCP servers like GitHub, Slack, and filesystem integrations. Organizations use it to build auditable, reusable data layers where every agent action carries a chain of custody — making AI-generated decisions trusted and traceable by humans and downstream agents alike.

Prerequisites

  • Docker (for running support containers: ArangoDB, Postgres, MinIO)
  • Git for cloning the repository if building from source
  • An MCP-compatible client such as Claude Desktop, VS Code, or Cursor
  • An OAuth2 provider account or credentials (Google, Microsoft Entra, Auth0, or custom OIDC) for authentication
  • An LLM API key (configured via the first-boot setup wizard)
1

Install Agience on your machine

Run the one-command installer appropriate for your OS. The installer sets up the full platform at https://home.agience.ai, which always resolves to 127.0.0.1 — traffic never leaves your machine. Caddy handles TLS automatically.

# macOS / Linux
curl -fsSL https://get.agience.ai/install.sh | sh

# Windows (PowerShell)
irm https://get.agience.ai/install.ps1 | iex
2

Start the platform

Use the agience CLI to bring up all containers. The first boot launches a setup wizard that walks you through OAuth provider configuration and LLM API key entry — no manual .env editing required.

agience up
3

Complete the first-boot setup wizard

Open https://home.agience.ai in your browser. The wizard will prompt for your OAuth2 provider details and an LLM API key. Once saved, Agience generates scoped API keys for each agent or MCP connection.

4

Locate your MCP gateway URL

After the platform is running, each registered MCP server is available at /{server_id}/mcp on the Chorus gateway. The platform advertises its MCP endpoints via /.well-known/mcp for auto-discovery by compatible clients.

# Gateway endpoint pattern
https://home.agience.ai/{server_id}/mcp

# Discovery endpoint
https://home.agience.ai/.well-known/mcp
5

Configure your MCP client

Add Agience to Claude Desktop or another MCP client using the gateway URL and the scoped API key generated in the setup wizard.

{
  "mcpServers": {
    "agience": {
      "url": "https://home.agience.ai/{server_id}/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_SCOPED_API_KEY"
      }
    }
  }
}
6

Manage the platform

Use the agience CLI for day-to-day operations — stopping, restarting, and pulling updates for all containers.

agience down    # Stop all containers
agience update  # Pull latest images and restart

Agience Core Examples

Client configuration (Claude Desktop)

Connect Claude Desktop to the local Agience MCP gateway using a scoped API key obtained from the Agience dashboard.

{
  "mcpServers": {
    "agience": {
      "url": "https://home.agience.ai/chorus/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_SCOPED_API_KEY"
      }
    }
  }
}

Prompts to try

Once connected, use these prompts to explore Agience's artifact and knowledge management capabilities.

- "Ingest this meeting transcript and extract decisions, actions, and constraints as typed artifacts."
- "Search the knowledge base for all decisions made about the Q3 project last month."
- "Show me the provenance chain for artifact ID abc-123 — who created it, from what input, and under whose authority?"
- "List all agent persona servers currently registered and their status."
- "Commit the draft artifacts in my workspace to the Q3 collection for review."

Troubleshooting Agience Core

The installer fails on a restricted network or TLS certificate errors appear

Use plain HTTP mode: run the installer with --mode plain (Linux/macOS) or -Mode plain (PowerShell). The platform will start at http://localhost:8080 without requiring a domain or certificate.

The MCP gateway returns 401 Unauthorized when connecting from Claude Desktop

Regenerate a scoped API key from the Agience dashboard, ensuring the key has read/write permissions for the server_id you are connecting to. Scoped keys are identity-bearing — each agent or client must have its own key.

Docker containers fail to start or ArangoDB exits immediately

Ensure Docker Desktop is running and has at least 4 GB of memory allocated. Run 'agience down' then 'agience up' again. Check logs with 'docker compose logs graph' for ArangoDB-specific errors.

Frequently Asked Questions about Agience Core

What is Agience Core?

Agience Core is a Model Context Protocol (MCP) server that operating system that ai workflows trust. structure messy inputs. establish identity. track provenance. build the durable data layer your ai workflows need to be trusted, reusable, and auditable. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Agience Core?

Follow the installation instructions on the Agience Core GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Agience Core?

Agience Core works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Agience Core free to use?

Yes, Agience Core is open source and available under the NOASSERTION license. You can use it freely in both personal and commercial projects.

Browse More Knowledge & Memory MCP Servers

Explore all knowledge & memory servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "agience-core": { "command": "npx", "args": ["-y", "agience-core"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Agience Core?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides