skills.rest
nebius
nebiusOFFICIAL
Explore all skills by this author →
WebsiteGitHub

agent-visual-feedback

Analyze Rerun canvases, video frames, and images to generate diagnostic feedback.

17|8|Updated Apr 7, 2026
Search Tags:#agent#multimodal#simulation#computer vision#robotics#rerun
One-click install
npx skills add https://github.com/nebius/nebius-physical-ai --skill agent-visual-feedback
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-visual-feedback
Source: https://github.com/nebius/nebius-physical-ai/tree/main/skills/atomic/agent-visual-feedback
Command: npx skills add https://github.com/nebius/nebius-physical-ai --skill agent-visual-feedback

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill solves the challenge of interpreting complex visual data from simulation and robotics environments, allowing the agent to provide immediate, context-aware feedback on Rerun frames, videos, and images.

Core Features & Use Cases

  • Visual Critique: Provides actionable feedback on simulation rollouts, skeleton movements, and environment meshes.
  • Multimodal Analysis: Interprets diverse inputs including Rerun canvases, video frames, and data-pane JSON.
  • Use Case: A robotics researcher can trigger a visual critique of a failed sim-to-real rollout to identify specific task progress cues or defects in the policy execution.

Quick Start

Use the agent-visual-feedback skill to describe the current viewer content and provide actionable critique.

Frequently Asked Questions about agent-visual-feedback

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I analyze visual data from a robotics simulation rollout to find execution defects?▼

To critique failed robotics rollouts visually, provide Rerun canvases, video frames, and data-pane JSON to generate structured diagnostic reports identifying task progress cues and policy execution defects.

How does multimodal visual critique work for interpreting simulation environments?▼

Multimodal visual critique operates by processing Rerun canvases and video frames through vision-language models to return structured diagnostic reports assessing environment meshes and skeleton movements.

Does Rerun canvas visualization support automated visual feedback for robotics research?▼

Yes, Rerun canvas visualization supports automated visual feedback by operating on image artifacts and video frames to deliver immediate, context-aware operator feedback for robotics simulation environments.

What do I need to generate actionable operator feedback from image artifacts in a sim-to-real environment?▼

Generating actionable operator feedback from image artifacts requires integrating vision-language models and the NPA agent chat queue to process visual context and return structured diagnostic reports.

What are the limitations of using automated visual analysis for robotics simulation data?▼

Automated visual analysis for robotics simulation data requires the NPA agent chat queue and vision-language models to process visual context, limiting standalone usage without these specific integrations.

Related Skills

Explore More Software Engineering Skills→
GregsGreyCodeGregsGreyCode
6

llava

Enable vision-language chat and image understanding with CLIP and Vicuna or LLaMA models.

Community
Advanced
LNC0831LNC0831
11

sim-bringup

Orchestrate behavioral RTL simulation pipelines with xvlog, xvhdl, and xelab.

Community
Advanced
nebiusnebius
17

onboard-world-model

Standardize containerization and multi-GPU deployment of action-conditioned world models.

Official
Advanced
zrongzrong
11

os-use

Automate cross-platform desktop tasks with screen capture and input simulation.

Community
Advanced