krusch-sequential-mcp
An advanced MCP server for reflective chain-of-thought with Semantic Plausibility Gating and DBOS PostgreSQL persistence.
README
<p align="center"> <strong>Krusch Sequential MCP</strong> </p>
<p align="center"> <strong>An advanced Model Context Protocol (MCP) server for reflective chain-of-thought, augmented with Semantic Plausibility Gating and DBOS PostgreSQL persistence.</strong> </p>
<p align="center"> <a href="https://www.npmjs.com/package/krusch-sequential-mcp"><img src="https://img.shields.io/npm/v/krusch-sequential-mcp.svg?style=flat-square" alt="NPM Version"></a> <a href="https://github.com/kruschdev/krusch-sequential-mcp/blob/main/LICENSE"><img src="https://img.shields.io/github/license/kruschdev/krusch-sequential-mcp.svg?style=flat-square" alt="License"></a> <img src="https://img.shields.io/badge/node-%3E%3D18-blue.svg?style=flat-square" alt="Node Version"> </p>
⚡ Why Krusch Sequential MCP?
The standard sequential-thinking MCP provides a great tool for chain-of-thought reasoning, but it suffers from the "Telephone Game" problem in multi-agent environments, where agents can confidently hallucinate ungrounded thoughts that poison the context window.
krusch-sequential-mcp solves this by introducing Semantic Plausibility Gating alongside a highly reliable DBOS PostgreSQL persistence layer.
Key Features
- 🧠 Semantic Plausibility Gating: Autonomously rejects drifted or hallucinated thoughts via an edge model evaluator.
- 💾 DBOS PostgreSQL Persistence: Synchronously persists every thought, branch, and revision into a
dbos_thoughtstable, creating an auditable DAG of reasoning. - 🛑 Deterministic State Reliability: Halts poisoned thought execution, forcing agents to re-evaluate their reasoning path.
- 🔌 Drop-In Replacement: Fully compatible with the standard
sequential-thinkinginterface while supporting the newgroundingContextparameter. - 📦 Zero External Dependencies: The plausibility evaluator is fully self-contained — no external toolkit required.
🧠 Architecture: Semantic Plausibility Gate
When an agent proposes a thought, the internal evaluator screens it against the provided groundingContext.
graph TD;
A[Agent Thought Proposed] --> B{Grounding Context Provided?};
B -- No --> C[Accept & Persist to DBOS];
B -- Yes --> D[Edge Model Evaluator];
D -- Plausible --> C;
D -- Hallucinated/Drifted --> E[Reject Thought];
E --> F[Return Soft Error to Agent];
F --> G[Agent Re-evaluates];
📦 Installation
npm install -g krusch-sequential-mcp
Or configure it in your MCP settings file (e.g., claude_desktop_config.json or .cursor/mcp.json):
{
"mcpServers": {
"krusch-sequential-mcp": {
"command": "npx",
"args": ["-y", "krusch-sequential-mcp"]
}
}
}
🚀 Quick Start Guide
Agents can invoke the sequentialthinking tool with the standard parameters (thought, thoughtNumber, totalThoughts, nextThoughtNeeded, etc.).
To engage the plausibility gate, include the groundingContext parameter in your tool call:
{
"thought": "Since the user is asking about the database schema, I will assume it uses MongoDB and write a query for it.",
"thoughtNumber": 1,
"totalThoughts": 3,
"nextThoughtNeeded": true,
"groundingContext": "The current codebase exclusively uses DBOS PostgreSQL for persistence. No NoSQL databases are present."
}
Because the thought conflicts with the groundingContext, the evaluator will autonomously reject it, returning an error to the agent to rethink its approach.
⚙️ Environment Variables
| Variable | Required | Default | Description |
|---|---|---|---|
DATABASE_URL |
No | (none — persistence disabled) | PostgreSQL connection string (e.g., postgres://user:pass@localhost:5432/mydb). If not set, the server runs in memory-only mode. |
OLLAMA_URL |
No | http://localhost:11434 |
Base URL to the Ollama service for plausibility checks. |
PLAUSIBILITY_MODEL |
No | qwen2.5-coder:1.5b |
Ollama model used for plausibility screening. Should be a small, fast model. |
Copy .env.example for a quick start:
cp .env.example .env
🤝 Contributing
We welcome contributions! Please ensure your tests pass and adhere to the project formatting standards.
Run tests via npm run build and npm start (or node build/index.js).
📄 License
MIT License © 2026 kruschdev
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
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
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.