9-axis-human-sonar-mcp

Model Context Protocol server for 9-axis conversation intelligence. Gives any MCP-aware LLM client (Claude Code, Claude Desktop, Cursor, etc.) tools to map transcripts across nine dimensions -- seven primes plus substrate (◊) and theta (θ).

Algorithms extracted verbatim from 9-Axis Human Sonar.

Install

npm install -g @ai-native-solutions/9-axis-human-sonar-mcp

Or run without installing:

npx @ai-native-solutions/9-axis-human-sonar-mcp

Related MCP server: fallcrm-elite-mcp

Wire into Claude Desktop

~/.config/claude/claude_desktop_config.json (macOS/Linux) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "9-axis-human-sonar": {
      "command": "npx",
      "args": ["-y", "@ai-native-solutions/9-axis-human-sonar-mcp"]
    }
  }
}

Restart Claude Desktop. The tools appear in the model's toolbox.

Wire into Claude Code

Add to ~/.claude.json:

{
  "mcpServers": {
    "9-axis-human-sonar": {
      "command": "npx",
      "args": ["-y", "@ai-native-solutions/9-axis-human-sonar-mcp"]
    }
  }
}

Tools

sonar_map_transcript

Map a transcript across the 9 axes.

{ "text": "COACH: ...\nCLIENT: ...", "focus": "CLIENT" }

Returns { axis_scores, thin, defended, strongest, weakest, turns_analysed, focus }.

sonar_next_probe

Suggest a probe targeting the weakest axis.

{ "current_map": { "substrate": {...}, "bloom": {...}, ... } }

Returns { axis, sym, name, probe }.

sonar_export

Schema-tag a map for storage.

Resources

• 9-axis://axes -- full axis definitions (prime, symbol, name, question, keyword sets)

• 9-axis://probes -- four probes per axis

Companion repos

• SDK: 9-axis-human-sonar-sdk -- programmatic API

• API: 9-axis-human-sonar-api -- HTTP server + Docker

• Tool: 9-axis-human-sonar -- shipping HTML app

License

MIT -- AI-Native Solutions.