Marmot provides a unified data catalog for discovering and managing data assets across an organisation. This skill exposes CLI, REST API and MCP server capabilities so an agent can search for datasets, retrieve asset details, view upstream/downstream lineage, query glossary terms, list pipeline runs, manage tags and owners, and fetch aggregated metrics. It supports both read and administrative workflows (with appropriate credentials).
Use this skill when a user asks to find or understand data (tables, topics, dashboards, APIs), wants ownership or lineage information, needs glossary definitions, or asks for metrics and run histories. Ideal for queries like "find a table for customer orders", "who owns this dataset?", "show lineage for sales_summary", or "top queries last 30 days". Prefer MCP endpoints when available for tighter agent integration; fall back to the CLI or REST API for unsupported operations.
Agents that can run shell commands or call HTTP APIs will benefit most (Copilot/Codex-style tools, Claude Code, Gemini CLI). MCP-aware agents that support MCP protocol integrations get the tightest experience.
This skill has not been reviewed by our automated audit pipeline yet.
