Language: 日本語English

rill Plugin

Author
Rill Data
Category
Development
Topics
Data Engineering & Analytics · Claude Code Customization & Workflow · AI Agents & AI App Development
First seen
2026-07-09
Last confirmed
2026-07-13
Explanation last updated
2026-07-12

The explanation below is AI-generated. Please verify it against the sources.

"rill" is a plugin that packages installable, modular skills for AI agents working with the Rill business intelligence platform. According to the GitHub README, it can be installed via npx skills add rilldata/agent-skills or as a Claude Code plugin, and it bundles skill files covering Rill development, canvas dashboards, connectors, explore dashboards, metrics views, models, the rill.yaml project file, and themes. The repository is synced from the output of rill init --agent agentsmd, the same instructions that power Rill's first-party coding assistant. Separately, the Rill homepage documents an ai_instructions configuration field that can be added to rill.yaml (project-wide) or to a metrics view YAML (view-specific) to give Rill's AI Chat and MCP Server additional context, such as business terminology, metric routing rules, and known data quirks. The MCP tools referenced in the README require a running Rill Developer server (rill start) on localhost:9009.

Overview

Rill is a business intelligence platform with project files (models, metrics views, dashboards), an AI Chat feature, and an MCP Server; the "rill" item is a set of AI-agent skills for developing and querying Rill projects, plus documentation on configuring AI context via ai_instructions.

What you can do with rill

  • Install the skills with npx skills add rilldata/agent-skills or as a Claude Code plugin (/plugin marketplace add rilldata/agent-skills then /plugin install rill@rilldata), per the GitHub README
  • Use bundled skill files for Rill development, canvas dashboards, connectors, explore dashboards, metrics views, models, rill.yaml, and themes, per the GitHub README
  • Run the MCP tools against a local Rill Developer server via rill start on localhost:9009, per the GitHub README
  • Add project-wide ai_instructions in rill.yaml to define business rules, terminology, and metrics-view routing, per the Rill homepage
  • Add metrics view-specific ai_instructions in a .yaml file to clarify metrics, dimensions, or data quirks for that view, per the Rill homepage
  • Rely on automatic inclusion of measure/dimension descriptions and metrics view metadata in the AI context without duplicating them in ai_instructions, per the Rill homepage
  • Add visualization-tip instructions to guide how AI clients (e.g., via the MCP Server) present data, such as chart or table formatting preferences, per the Rill homepage

Sources

Original description (English)

Skills for developing and querying projects in the Rill business intelligence platform

History of rill

Back to list