Classifer
Enables Claude to operate and reconfigure a Discord server through a registry-based system, using 60 REST tools to manage actions, components, scheduling, and safety without code changes or bot restarts.
README
<div align="center">
Cognition
A Discord system where behavior is stored as data and applied live.
</div>
Overview
Cognition separates Discord behavior from the bot source code. Actions, components, triggers and schedules live in a SQLite registry. The running bot reads that registry when an interaction happens and executes the matching behavior.
Claude can inspect and change the registry through the MCP server. Most server changes can be applied without adding a new handler or restarting the Discord process.
Architecture
Cognition runs as two processes.
flowchart LR
Claude[Claude] --> MCP[MCP server]
MCP <--> Registry[(SQLite Registry)]
MCP --> Discord[Discord REST API]
Discord <--> Bot[Cognition bot]
Bot <--> Registry
MCP server
Used for reading server state, editing registry data, scheduling work and performing controlled structural operations through Discord REST.
Cognition bot
Keeps the Discord gateway connection alive and handles interactions, events and scheduled execution.
Registry
Stores the behavior the bot can execute:
- actions
- components
- triggers
- schedules
- sessions
- counters
- snapshots
- audit records
Registry actions
Actions are small primitives that can be composed into larger flows.
{
"key": "ticket_open",
"kind": "channel_create",
"params": {
"name": "ticket-{{user.name}}",
"parent_id": "CATEGORY_ID"
}
}
The dispatcher looks up the action when the component is used and passes it to the executor.
Nested actions can use values created earlier in the same flow such as {{created.id}}.
Components and events
A registry action can be started by:
| Source | Example |
|---|---|
| Component | Button or modal interaction |
| Trigger | Member join, message or reaction event |
| Schedule | Cron based execution |
All three use the same registry and executor.
Safety
Structural changes create snapshots before writing where restoration is possible.
Destructive operations use a separate confirmation flow. The system first prepares a plan tied to the exact operation. The final apply step only accepts the matching confirmation token.
Unknown predicates and invalid requirements fail closed instead of allowing the action to continue.
Setup
Requirements:
- Node.js 22.5 or newer
- Discord application
- Discord bot with the required intents
npm install
cp .env.example .env
Set the Discord token and guild ID in .env, then run:
npm run smoke
npm run bootstrap
npm run bot
Install the Claude Code plugin with:
node scripts/install-plugin.js
The MCP tools can also be called directly from the command line:
npm run call guild_snapshot
npm run check
Project layout
bot/ Discord gateway process
shared/ Registry and execution logic
classifer/ MCP server
skills/ Operating guidance for Cognition tools
scripts/ Setup and project utilities
License
This repository is source available. See LICENSE for the project terms.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.