Happy slice of mcp pizzaMCP.pizza

mcp-server-mas-sequential-thinking

MCP.Pizza Chef: FradSer

Every question you send is examined from six angles at once — the facts, the risks, the upside, the gut reaction, the creative alternatives, and a check on the reasoning itself — then pulled together into a single answer. It suits decisions you want stress-tested rather than quick lookups. Getting it running is a job for someone technical: it installs through Python, and it needs a paid account key from DeepSeek, Groq, OpenRouter, Anthropic, or GitHub Models, because it calls those services on every step.

Other
Writing

Use This MCP server To

Pressure-test a decision before I commit to it List the weak points in a plan I wrote Get the case for and against the same idea Work through a knotty problem one step at a time Check my own reasoning for blind spots

README

Sequential Thinking Multi-Agent System (MAS)

Python Version Framework Twitter Follow

English | 简体中文

An MCP server that processes sequential thoughts through a team of specialized AI agents, each analyzing the problem from a different cognitive perspective.

What This Is

This is an MCP server, not a standalone application. It runs as a background service that extends an MCP-compatible LLM client (like Claude Desktop) with structured sequential-thinking capabilities. It exposes one tool, sequentialthinking, that runs every thought through a fixed multi-agent workflow: an initial synthesis, several specialist agents thinking in parallel, and a final synthesis that answers the original question.

How It Works

The system uses a fixed full_exploration strategy for every request. The AI complexity analyzer still runs to record diagnostic metadata (complexity score, problem type, required thinking modes), but it no longer changes the execution path — all thoughts take the same route:

flowchart TD
    A[Input Thought] --> B[AI Complexity Analyzer]
    B --> C[Complexity Metadata Stored]
    C --> D[Fixed Strategy: full_exploration]
    D --> E[Step 1: Initial Synthesis]
    E --> F[Step 2: Parallel Specialist Agents]
    F --> G[Step 3: Final Synthesis]
    G --> H[Unified Response]
Loading

mcp-server-mas-sequential-thinking FAQ

Does this cost money to run?
Usually yes. It needs a key from an outside model provider such as DeepSeek, Groq, OpenRouter, Anthropic, or GitHub Models, and each question runs several passes. Running models locally through Ollama is supported as a free alternative.
Which apps does it work in?
Any app that supports MCP. Claude Desktop is the worked example in the instructions.
How hard is the setup?
You need Python 3.10 or newer, you install it from the source code, and you set a provider key, so it is aimed at people comfortable in a terminal.
Can I use this to talk through a business decision?
Yes — that is what it is built for. It runs your question past several perspectives and then combines them into one recommendation.
Is it slower than just asking normally?
Yes, noticeably. Each request runs seven separate thinking passes, some allowed two to four minutes, so expect to wait.
Can it look things up on the web?
Only if you add a separate Exa key. Without one it reasons from what the model already knows.
Is it still maintained?
Yes, with commits within the last few days, though the author still labels the project a work in progress.