DataRobot for Developers — integrating with the Google Antigravity CLI

2026-06-15 · Source: Blog | DataRobot · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Advanced, short

Summary

DataRobot has released a new plugin for Google's Antigravity CLI, the latest agentic coding interface replacing the Gemini CLI. This plugin, available from the datarobot-oss/datarobot-agent-skills open-source repository, integrates DataRobot's full skill set, including datarobot-setup and datarobot-agent-assist, into Antigravity's agent and slash-command interface. A key feature is the "/datarobot-skills:datarobot-external-agent-monitoring" skill, designed to provide production-grade tracing for debugging agent behavior. This skill provisions a new DataRobot Use Case, instruments agents to emit traces via OpenTelemetry, and generates a "monitoring_setup.md" artifact. The DataRobot tracing interface then visualizes detailed request histories, showing end-to-end latency, token consumption, and complete span trees, which proved crucial in diagnosing an example LangGraph agent's failure to access calendar data. The DataRobot CLI also offers real-time trace updates for local development.

Key takeaway

For AI Engineers debugging complex LLM agents or preparing them for production, integrating the DataRobot plugin with Google's Antigravity CLI offers crucial operational visibility. You should install the plugin and utilize the "/datarobot-skills:datarobot-external-agent-monitoring" skill to instrument your agents with OpenTelemetry. This provides detailed trace data, revealing hidden failures and performance bottlenecks that conversational outputs alone cannot expose, ensuring your agents behave as expected at scale.

Key insights

Debugging LLM agents effectively demands structured trace data to reveal hidden failures and operational gaps.

Principles

• Agent failures are often masked by plausible LLM responses.
• Structured tracing is essential for agent observability.
• OpenTelemetry enables standardized agent instrumentation.

Method

Install the DataRobot plugin, invoke "/datarobot-skills:datarobot-external-agent-monitoring" to instrument an agent with OpenTelemetry, then run the agent and analyze traces in DataRobot's interface or CLI.

In practice

• Install DataRobot plugin for Antigravity CLI.
• Use "/datarobot-skills:datarobot-external-agent-monitoring" for agent tracing.
• Monitor local agent traces with "dr xp --entity-id=".

Topics

• Antigravity CLI
• LLM Agents
• Agent Debugging
• OpenTelemetry
• DataRobot
• Observability

Code references

• datarobot-oss/datarobot-agent-skills
• datarobot-oss/datarobot-agent-skills

Best for: AI Engineer, Machine Learning Engineer, MLOps Engineer

Related on AIssential

• How to Find the Agent Failures Your Evals Miss with Scott Clark - #767 — Effective LLM operations require a hierarchy of observability, with analytics crucial for discovering unknown issues and driving iterative improvement.
• The DataRobot platform as skills in Claude Code — Packaging platform operational judgment as agent skills enables reliable, governed AI agent production deployments.
• Stop Celebrating The Illusion of Progress In AntiGravity Update! — Celebrating basic operational transparency as progress highlights fundamental platform design deficiencies.
• Agent Tracing and Observability: Log & Debug Complex AI Systems — Multi-agent systems require specialized observability beyond traditional logging to debug coordination and self-modification failures.
• LLM Observability, Evaluation, Experimentation Platform — Dat Ngo, Arize — LLM application development demands structured observability, evaluation, and automated experimentation to manage non-deterministic behavior and ensure continuous improvement.
• How agent o11y differs from traditional o11y — Phil Hetzel, Braintrust — Agent observability requires specialized approaches due to LLM non-determinism, complex trace data, and diverse stakeholder involvement.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Blog | DataRobot.