jellyfish-mcp is a Model Context Protocol server designed to retrieve and analyze data from Jellyfish's API. It enables hosts such as Claude Desktop or Cursor to interact with a Jellyfish instance through natural language queries about engineering metrics, team data, allocations, deliveries, and other information exposed by the Jellyfish API. The MCP server exposes a set of tools that map to Jellyfish API endpoints: general utilities (get_api_schema, list_endpoints, get_endpoint); allocations (allocations_by_person, allocations_by_team, allocations_by_investment_category, allocations_by_investment_category_person, allocations_by_investment_category_team, allocations_by_work_category, allocations_by_work_category_person, allocations_by_work_category_team, allocations_filter_fields, allocations_summary_by_investment_category, allocations_summary_by_work_category); delivery (deliverable_details, deliverable_scope_and_effort_history, work_categories, work_category_contents); metrics (company_metrics, person_metrics, team_metrics, team_sprint_summary, unlinked_pull_requests); people (list_engineers, search_people); teams (list_teams, search_teams). Each tool corresponds to a specific Jellyfish API endpoint and allows retrieval or search of data as described in the API. Setup supports Desktop Extension for Claude Desktop or local setup; credentials can be stored via keyring or environment variables; optional PromptGuard 2 integration provides prompt-injection mitigation. The README also describes configuration steps for VSCode, Cursor, Claude Desktop, and troubleshooting guidelines. This MCP server enables natural language access to discover API schemas, list endpoints, and query metrics and organizational data across Jellyfish.