karst MCP Server

by moin105 Your server? Claim it

local only

karst runs on your own machine — the client starts it, so there is no endpoint to ping. 198 installs a week from pypi. Last commit 27 Jul 2026.

Code context for AI dev tools: pack-scoped, cited code retrieval over MCP.

Installs per day peak 84 · avg 36 · -26% w/w

a month agotoday

198

Installs / week

pypi · karst

Stars

0 open issues

27 Jul 2026

Last commit

0 releases in 90 days

Apache-2.0

License

Python

What the code does

We read the source, 7 d ago · rules 3dff92dd89df

A tool parameter reaches a dangerous call

A value the model can set ends up inside a file or shell call. That is not a flaw by itself — for a terminal server it is the job — but it is where things go wrong when it is not.

A tool parameter reaches a dangerous call search_code.repo_path → karst/graph_cli.py:191, search_code.repo_path → karst/review_cli.py:175

    proc = subprocess.run(

Is this your server and something here is wrong? Tell us — corrections are free and do not require a plan.

A tool parameter here reaches a dangerous call

That is not a flaw by itself — but it is where things go wrong when it is not the job. We re-read this code on every release. Watch it and you hear from us the day another one appears.

Connect this server

This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.

Claude Code Claude Desktop Codex Cursor VS Code

run in your terminal

claude mcp add karst -- uvx karst

~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "karst": {
      "args": [
        "karst"
      ],
      "command": "uvx"
    }
  }
}

~/.codex/config.toml

[mcp_servers.karst]
command = "uvx"
args = ["karst"]

.cursor/mcp.json

{
  "mcpServers": {
    "karst": {
      "args": [
        "karst"
      ],
      "command": "uvx"
    }
  }
}

.vscode/mcp.json

{
  "mcpServers": {
    "karst": {
      "args": [
        "karst"
      ],
      "command": "uvx"
    }
  }
}

Alternatives to karst

same job, measured the same way

Code Context Gate

by wisdomrock

Context-aware code retrieval MCP server — ranked, gated results for AI coding agents

27 installs/wk local only

ContextStream MCP Server

by contextstreamio

ContextStream MCP server for code context, memory, search, and AI tools.

1 106 installs/wk local only

Skeletongraph

by yashdoke7

Zero-LLM structural code retrieval for AI coding agents, served over MCP.

28 installs/wk local only

Cgraphy

by pmgarg

Code knowledge graph over MCP: token-budgeted context for AI coding agents.

63 installs/wk local only

RTFM

by roomi-fields

The open retrieval layer for AI agents — index code, docs, data. Search via MCP.

9 734 installs/wk local only

Silica Core

by kiycoh

Retrieval over the folder the session opened: files, search, read, code_pack, write_note.

local only

Slimdex MCP

by siddhukaushik

Narrow code retrieval for agents: outlines, symbol context, dep graph, persistent memory.

37 installs/wk local only

LeanToken

by morluto

Token-bounded repository context for coding agents over MCP.

local only

karst — questions

Answers built from our own checks of this server.

Why is there no uptime for karst?

karst runs on your own machine over stdio — there is no network address to reach, so uptime cannot be measured for it by anyone. What can be measured is adoption: the pypi package karst was installed 198 times last week.

How do I connect karst?

Copy the ready config from this page — we generate it for Claude Code, Claude Desktop, Codex, Cursor and VS Code, each with the file path that client actually reads. It runs locally, so the command pulls karst straight from pypi; nothing to host, nothing to sign up for.

How many people use karst?

The pypi package karst was installed 198 times in the last week. Week over week that is -26%. We show installs rather than GitHub stars on purpose: a star is a bookmark, an install is someone actually running it.

Is karst open source?

Yes — it is published under the Apache-2.0 licence, written in Python and 0 stars on GitHub. The source link is on this page, so you can read exactly what it does with your data before you connect it.