analysis-mcp

A FastMCP server for critical thinking and multi-perspective analysis of current affairs.

Uses the LLM-Orchestrator Pattern: Tools return structured prompts for the calling LLM to execute, ensuring provider-agnostic operation and full transparency.

🧠 How It Works

You ask the LLM to analyze something
LLM calls MCP tool to get a structured analysis plan
Tool returns {trace_id, outline, next_prompt}
LLM executes the next_prompt using its own model
Result combines tool planning + LLM generation

This pattern means:

✅ Works with any LLM provider (Claude, GPT, local models)
✅ No API keys needed in the MCP server
✅ Full traceability via trace_id
✅ Composable with other MCP tools

Features

Cognitive Tools:

deconstruct_claim - Plan a claim deconstruction (returns structured prompt)
compare_positions - Plan multi-perspective analysis (returns structured prompt)
apply_lens - Plan lens-based analysis through 9 analytical frameworks
get_trace - Retrieve previous analysis plans for iteration

9 Analytical Lenses: historical, economic, geopolitical, psychological, technological, sociocultural, philosophical, systems, media

Quick Start with Claude Desktop

Install via uvx (recommended):

Edit your Claude Desktop config file:

MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json

Add this to the mcpServers section:

{ "mcpServers": { "analysis-mcp": { "command": "uvx", "args": [ "git+https://github.com/YOUR_USERNAME/analysis_mcp", "analysis-mcp" ] } } }

Restart Claude Desktop
Verify installation: Look for the 🔌 icon in Claude Desktop showing the analysis-mcp server is connected

Alternative: Local Development Installation

If you want to modify the code or run it locally:

# Clone the repo git clone https://github.com/YOUR_USERNAME/analysis_mcp.git cd analysis_mcp # Create virtual environment python -m venv .venv source .venv/bin/activate # On Windows: .venv\Scripts\activate # Install in editable mode pip install -e ".[dev]" # Run tests pytest -v # Run server directly (for testing) python -m analysis_mcp.server

For local development in Claude Desktop, update your config to point to the local path:

{ "mcpServers": { "analysis-mcp": { "command": "python", "args": [ "-m", "analysis_mcp.server" ], "cwd": "/absolute/path/to/analysis_mcp", "env": { "PYTHONPATH": "/absolute/path/to/analysis_mcp/src" } } } }

Usage Examples

Once connected to Claude Desktop, you can use these tools:

Example 1: Deconstruct a claim

Analyze this claim: "AI will replace all human jobs within 10 years"

The tool returns:

{ "trace_id": "deconstruct-1730830000000-abc123", "outline": { "claim": "AI will replace all human jobs within 10 years", "analysis_sections": ["assumptions", "evidence", "implications", "hidden_premises"] }, "next_prompt": "TASK: Deconstruct the following claim...\n\nCONTENT:\nAI will replace all human jobs within 10 years\n\nINSTRUCTIONS:\n1. List all implicit and explicit ASSUMPTIONS..." }

Claude then executes the next_prompt and provides the full analysis.

Example 2: Compare perspectives

Compare progressive vs conservative perspectives on universal basic income

Example 3: Apply analytical lens

Apply an economic lens to analyze: "Federal Reserve raises interest rates"

Example 4: Recall previous analysis

Retrieve trace deconstruct-1730830000000-abc123

🔄 LLM-Orchestrator Pattern Details

Each tool is a prompt compiler, not a generator:

User → LLM: "Analyze X" LLM → MCP Tool: deconstruct_claim("X") Tool → LLM: {trace_id, outline, next_prompt} LLM → LLM: executes next_prompt LLM → User: final analysis

Why this pattern?

Server never calls external APIs (no cost, no keys)
Uses the calling LLM's provider & model
Deterministic planning + flexible generation
Easy to trace, log, and iterate
Composable with other tools

Available Lenses

historical - Compare to precedents and patterns
economic - Analyze resource flows and incentives
geopolitical - Examine power balances and strategy
psychological - Identify biases and manipulation
technological - Explore tech's role and impact
sociocultural - Analyze identity and narratives
philosophical - Apply ethical frameworks
systems - Map feedback loops and leverage points
media - Deconstruct framing and agenda-setting

Trace Storage

Analysis plans are logged to ~/.analysis_mcp/traces/ as JSON files. Each trace contains:

trace_id - Unique identifier
tool - Which tool was called
input - Original parameters
outline - Structured analysis plan
next_prompt - The prompt for LLM execution
timestamp - When it was created

Use get_trace(trace_id) to retrieve any previous analysis plan.

Troubleshooting

Server not connecting?

Verify uvx is installed: pip install uvx
Check Claude Desktop logs (Help → View Logs)
Ensure your config JSON is valid

Tools not appearing?

Restart Claude Desktop after config changes
Check the 🔌 icon shows "analysis-mcp" as connected

Contributing

Pull requests welcome! Please run tests before submitting:

pytest -v

This server cannot be installed

-

security - not tested

F

license - not found

-

quality - not tested

How are these scores calculated?

local-only server

The server can only run on the client's local machine because it depends on local resources.

Provides cognitive tools for critical thinking and multi-perspective analysis of current affairs through structured prompts, including claim deconstruction, perspective comparison, and analysis through 9 analytical lenses (historical, economic, geopolitical, etc.).

Analysis MCP