AI Integration

GeoHammer can hand the data you are working on to an AI assistant. It does this through the MCP server built into the application, which assistants connect to as a tool provider.

The server is off by default. To turn it on, open Settings → AI Integration, check Enable MCP server and press Save. It listens on the loopback interface only, at:

http://localhost:41693/mcp

Use the copy button next to the MCP server address to put it on the clipboard. The setting is remembered between runs, but the server itself is part of GeoHammer: it starts with the application and stops when you close it.

Once connected, an assistant can list the files you have open, read and write data series, run filters and gridding, cut survey lines, place marks and take screenshots of the charts. All of it works on the in-memory data of the files open in GeoHammer. Nothing is written to disk unless you save the file yourself.

Claude Code CLI

1. Install Claude Code by following the Claude Code installation guide.

macOS, Linux:

curl -fsSL https://claude.ai/install.sh | bash

Windows PowerShell:

irm https://claude.ai/install.ps1 | iex

2. Register GeoHammer MCP server. In Settings → AI Integration, press Enable in the Claude Code Integration row, or run:

claude mcp add -t http -s user geohammer http://localhost:41693/mcp

Changes take effect in new Claude Code sessions.

Codex CLI

1. Install Codex by following the Codex CLI installation guide.

macOS, Linux:

curl -fsSL https://chatgpt.com/codex/install.sh | sh

Windows PowerShell:

irm https://chatgpt.com/codex/install.ps1 | iex

2. Register GeoHammer MCP server. In Settings → AI Integration, press Enable in the Codex Integration row, or run:

codex mcp add geohammer --url http://localhost:41693/mcp

Changes take effect in new Codex sessions. On older Codex versions, HTTP servers are behind a flag: if geohammer is registered but its tools never appear, add experimental_use_rmcp_client = true under [features] in ~/.codex/config.toml, or update Codex.