ADE Adds Visual Comments for Agent Work

// 1h agoPRODUCT UPDATE

ADE Adds Visual Comments for Agent Work

ADE now lets developers comment directly on image and video artifacts, then send those notes to an agent with the next message. The update makes visual feedback a first-class part of agentic development workflows, complementing ADE’s screenshot and video proof features.

// ANALYSIS

This is a small UX change with outsized value: visual feedback is faster and less ambiguous than describing screenshots in prose.

–Helps agents interpret UI bugs, layout issues, simulator states, and failed test evidence
–Keeps annotations attached to the exact visual artifact under review
–Shortens the loop between agent-generated proof and human correction
–Makes ADE’s evidence-driven workflow more practical for frontend and app development
–Developers still need structured reproduction steps for complex or nonvisual failures

// TAGS

adeagentcoding-agentai-codingmultimodalvisiondevtool

DISCOVERED

1h ago

2026-11-03

PUBLISHED

1h ago

2026-11-03

RELEVANCE

8/ 10

AUTHOR

SharmaKarma1028

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

TUTORIAL35m ago

Entire Trails Run Custom Reviews on Every Push

Entire’s Trails attach agent sessions, checkpoints, custom review scores, findings, CI, and approvals to every code change. This guide shows how to enable presets like Slop Cop and build domain-specific agents for risks such as unsafe database migrations. [Source](https://entire.com/blog/how-to-run-custom-ai-code-evals-on-every-push-with-trails)

UPDATE39m ago

Priors Passes 1,000 User Rankings

Priors lets people rank AI models through quick head-to-head comparisons, then finds benchmark mixes that best match their preferences. It now reports more than 1,000 rankings and publishes shareable, evidence-backed leaderboards.

OPEN SOURCE40m ago

NVIDIA Auto Ontology Grounds Agent SQL

NVIDIA Auto Ontology is an Apache-2.0 Python framework that builds a semantic, graph-based context layer from enterprise metadata, helping agents map business terms to database fields, validate queries, and return provenance. It exposes this context through MCP, CLI, and API integrations.