MC Agent Toolkit

Monte Carlo's official toolkit for AI coding agents. Brings data observability — lineage, monitoring, validation, alerting, and metadata ingestion — directly into your development workflow. The toolkit bundles multiple skills into a single plugin that works across supported editors.

Features

The toolkit bundles the following capabilities as a single mc-agent-toolkit plugin. Each feature is a skill that can also be used standalone.

Skills are grouped by the job they help you do. Orchestrated workflows sequence individual skills into guided multi-step flows; atomic skills can be invoked directly by name. Both are loaded the same way.

Trust — pre-query and pre-build checks

Skill	Description	Details
Asset Health	Single-table health report: freshness, active alerts, monitor coverage, importance, and upstream issues. Run before building on a table.	README

Incident Response — triage, investigate, fix

Skill	Description	Details
Incident Response (workflow)	Orchestrates full incident lifecycle — triage → root cause → remediation → prevent recurrence.	SKILL
Automated Triage	Scores and prioritizes active alerts; runs deep troubleshooting on high-signal ones.	SKILL
Analyze Root Cause	Investigates incidents via lineage tracing, ETL checks, query analysis, and data profiling.	README
Remediation	Proposes and executes fixes for data-quality alerts; assesses blast radius before acting, or escalates with full context.	README

Monitoring — coverage gaps, monitor creation, noise reduction

Skill	Description	Details
Proactive Monitoring (workflow)	Sequences coverage analysis → gap identification → monitor creation into a guided flow.	SKILL
Monitoring Advisor	Identifies coverage gaps and creates monitors for warehouse tables or AI agents — validates tables and fields against your live workspace, emits monitors-as-code YAML.	README
Tune Monitor	Recommends sensitivity, segment, and schedule changes to reduce alert noise on an existing metric monitor.	SKILL

Prevent — catch issues before they ship

Skill	Description	Details
Prevent	Edit-lifecycle safety net for dbt/SQL: surfaces blast radius and monitor gaps before edits, generates monitors-as-code for new logic. Auto-activates via hooks.	README
Generate Validation Notebook	Generates targeted SQL validation queries for a dbt PR or local repo change.	README

Optimize — cost and performance

Skill	Description	Details
Storage Cost Analysis	Identifies storage waste (unread, zombie, dead-end tables); uses lineage to verify cleanup is safe and estimates savings.	README
Performance Diagnosis	Diagnoses slow pipelines and expensive queries across Airflow, dbt, Databricks, and other platforms.	README

Setup — ingestion and connections

Skill	Description	Details
Push Ingestion	Generates collection scripts to push metadata, lineage, or query logs to Monte Carlo from any data source.	README
Connection Auth Rules	Builds Connection Auth Rules JSON for a Monte Carlo connection type using live connector schemas.	SKILL

Installing the plugin (recommended)

Monte Carlo recommends installing the mc-agent-toolkit plugin. The plugin bundles all skills together with hooks, the Monte Carlo MCP server, and agent-specific capabilities — no separate MCP configuration or authentication setup needed. See the plugins page for the full list of supported coding agents.

Claude Code

1. Add the marketplace:

   /plugin marketplace add monte-carlo-data/mc-agent-toolkit
   

2. Install the plugin:

   /plugin install mc-agent-toolkit@mc-marketplace
   

3. Updates — claude plugin update pulls in the latest skill and hook changes.

See the Claude Code plugin README for detailed setup and usage.

For other coding agents (Cursor, Copilot CLI, OpenCode, Codex), see the plugins page for installation guides.

Using skills directly (advanced)

Skills can also be used standalone without the plugin. This is for users who want to install individual skills via registries or use them with agents not listed above.

Prerequisites

• A Monte Carlo account with Editor role or above

• Monte Carlo MCP server — configure with:

  claude mcp add --transport http monte-carlo-mcp https://mcp.getmontecarlo.com/mcp
  

  Then authenticate: run /mcp in your editor, select monte-carlo-mcp, and complete the OAuth flow.

  See official docs for other MCP clients and advanced options.

  Legacy: header-based auth (for MCP clients without HTTP transport)

  If your MCP client doesn't support HTTP transport, use .mcp.json.example with npx mcp-remote and header-based authentication. See the MCP server docs for details.

Installation

npx skills add monte-carlo-data/mc-agent-toolkit --skill prevent

Or copy directly:

cp -r skills/prevent ~/.claude/skills/prevent

See the skills directory for the full list and individual READMEs.

Telemetry

The Claude Code plugin sends anonymous skill-usage telemetry. No prompts or skill arguments are sent. To disable, set MC_AGENT_TOOLKIT_TELEMETRY_DISABLED=1. See the plugin README for details.

Contributing

See CONTRIBUTING.md for guidelines on adding skills, creating plugins, and submitting pull requests. It also covers plugin architecture and releasing new versions.

License

This project is licensed under the Apache-2.0 license — see LICENSE for details.

Security

See SECURITY.md for reporting vulnerabilities.