Data Olympus — Claude Code MCP server: features and setup
Data Olympus gives agents a Git-backed collection of engineering standards and decisions to consult while working on a project. Use it when existing team choices should guide new work.
claude mcp add --env KB_MAIN_PATH=<KB_MAIN_PATH> --env KB_INDEX_PATH=<KB_INDEX_PATH> --env KB_HTTP_PORT=<KB_HTTP_PORT> --transport stdio data-olympus -- uvx data-olympusWhat Data Olympus does
The knowledge base is made of Markdown files with YAML frontmatter and lives in Git. Its MCP server lets an agent look up a relevant standard or architectural decision; the CLI can create a bundle, lint its contents, and build an index. It targets decision lookup, not code search or finding where a symbol is used.
The format works with Open Knowledge Format v0.2 consumers and adds governed document fields and links between superseded decisions. Proposed edits pass through a queue before commit, while the MCP write pipeline uses a single writer. Python 3.13+ and uv are required; the quickstart commands depend on both.
What it helps with
- look up an accepted architectural decision
- check a library choice against team standards
- create a Markdown knowledge base
- lint and index knowledge-base documents
Example prompts for the agent
- Check the knowledge base for an accepted decision about choosing a library for this project.
- Find the standard that should guide this database migration.
- Create a knowledge base in
my-kband lint its documents.
How to install Data Olympus
Platform CLI
claude mcp add --env KB_MAIN_PATH=<KB_MAIN_PATH> --env KB_INDEX_PATH=<KB_INDEX_PATH> --env KB_HTTP_PORT=<KB_HTTP_PORT> --transport stdio data-olympus -- uvx data-olympusReplace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from data-olympus <command>
Platform CLI
codex mcp add data-olympus --env KB_MAIN_PATH=<KB_MAIN_PATH> --env KB_INDEX_PATH=<KB_INDEX_PATH> --env KB_HTTP_PORT=<KB_HTTP_PORT> -- uvx data-olympusReplace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from data-olympus <command>
Config file
{
"mcpServers": {
"data-olympus": {
"command": "uvx",
"args": [
"data-olympus"
],
"env": {
"KB_MAIN_PATH": "<KB_MAIN_PATH>",
"KB_INDEX_PATH": "<KB_INDEX_PATH>",
"KB_HTTP_PORT": "<KB_HTTP_PORT>"
}
}
}
}Replace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from data-olympus <command>
Config file
{
"servers": {
"data-olympus": {
"type": "stdio",
"command": "uvx",
"args": [
"data-olympus"
],
"env": {
"KB_MAIN_PATH": "<KB_MAIN_PATH>",
"KB_INDEX_PATH": "<KB_INDEX_PATH>",
"KB_HTTP_PORT": "<KB_HTTP_PORT>"
}
}
}
}Replace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from data-olympus <command>
Platform CLI
gemini mcp add -e KB_MAIN_PATH=<KB_MAIN_PATH> -e KB_INDEX_PATH=<KB_INDEX_PATH> -e KB_HTTP_PORT=<KB_HTTP_PORT> data-olympus uvx data-olympusReplace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from data-olympus <command>
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about Data Olympus
How do I install Data Olympus in Claude Code?
Run claude mcp add data-olympus -- uvx data-olympus. To set the paths and port, run `claude mcp add --env KB_MAIN_PATH=<KB_MAIN_PATH> --env KB_INDEX_PATH=<KB_INDEX_PATH> --env KB_HTTP_PORT=<KB_HTTP_PORT> --transport stdio data-olympus -- uvx data-olympus`.
What does Data Olympus do?
It stores project knowledge, engineering standards, and architectural decisions as Markdown files in Git. Agents can query the collection for guidance relevant to a coding decision.
Which prerequisites does Data Olympus have?
It requires Python 3.13+ and uv. Install the CLI with uv tool install data-olympus.
