Foxglove
Agentic data platform for robotics with visualization and debugging.
About
Foxglove is a visualization and debugging platform for robotics data, supporting live and recorded streams across MCAP, ROS 1, ROS 2, Protobuf, and JSON formats. Its modular Studio interface packs 3D views, image feeds, plots, diagnostics, and point-cloud panels into shareable layouts, with an extensions API for custom panels.
In 2026 the company launched an agentic layer: an Agent Sidebar inside the app plus a Foxglove MCP server so external AI agents can search, visualize, compare, debug, and curate multimodal datasets in natural language. Semantic Search, built with NVIDIA Cosmos world models, retrieves matching multimodal segments from large unlabeled corpora, alongside Comparison Mode and Remote Access for field debugging. The platform runs locally or in the cloud with team collaboration features.
What it can do
- Multimodal visualization panels for 3D, image, and sensor data
- MCAP-based logging, playback, and timeline review
- AI agent sidebar and MCP server for natural-language data work
- Semantic Search powered by NVIDIA Cosmos models
- Extensions API and registry for custom panels
Setup
Download Foxglove Studio for Linux, Windows, or macOS, or use the cloud platform.
Prerequisites
Robotics log data in MCAP, ROS bag, or a supported format.
Use cases
Works with
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 135; Trending score 82/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.