OMDS (oh-my-datascience) — data-science guardrails for coding agents: train/test leakage and metric-misuse checks, code-graph lineage, an experiment ledger and episodic memory, as portable Agent Skills + CLIs, an optional MCP server, and a Claude Code plugin. Includes E-GDS, an offline GEPA pipeline optimizer with gated promotion.

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

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

Or install with pip: pip install egds

README Excerpt

One Python package (`omds`) that does two things over the same cores: 1. **A data-science guardrail toolkit for coding agents.** A set of `omds-*` CLIs plus portable [Agent Skills](skills/) that let any coding agent check Python for train/test leakage and metric misuse, trace data lineage through a code graph, and keep a durable experiment ledger and episodic memory.

Tools (10)

codegraph_blamecodegraph_contextcodegraph_lineagedstools_ledger_showguardrails_checkmemory_addmemory_recallsessionpad_appendsessionpad_showsurface_references

Topics

agent-skillsclaude-codecode-graphdata-leakagedata-sciencegepallm-toolsmachine-learningmcppython

egds

Quick Install

README Excerpt

Tools (10)

Topics

Related AI/LLM Servers