3 items
Figma-style comments for localhost. Annotate DOM elements to generate high-context UI/UX instructions for your AI coding agents. Taco lets you leave Figma-style comments on your localhost web app and sends them straight to your AI coding agent โ Cursor, Claude Code, Windsurf, Antigravity, or Codex. NEW in v1.1.1: Taco now works with all five agents โ Cursor, Claude Code, Windsurf, Antigravity, and Codex (ChatGPT). You asked for Claude and Cursor support, it's here! Instead of manually describing UI bugs or layout shifts to your AI agent, you click the element and type what's wrong. Taco captures your comment along with precise element context (e.g. CSS selectors like "div#root > div:nth-child"), page metadata (e.g. URL link), and a screenshot of the element, and batches it into structured instructions. These batched instructions are delivered to your agent over MCP (Model Context Protocol) for your AI-driven implementation. Your agent finds the right component, makes the fix, and marks the comment resolved. Pins update live in your browser. Ever asked AI to tweak one button and it restyled your whole page? This won't happen with Taco because Taco pins your comment to the exact element, so your agent changes that element, and nothing else. HOW IT WORKS 1. Pick your agent (Cursor, Claude Code, Windsurf, Antigravity, or Codex) and run one setup command 2. Click any element on your localhost page and leave a comment, like in Figma 3. Your agent picks up the comment โ with selector, screenshot, and URL context โ and fixes it 4. Watch the pin turn resolved WHO IT'S FOR Developers building web apps with AI coding agents and assistants. If you've ever typed "the third button in the navbar is misaligned, no, the OTHER button" into a chat box, Taco is for you. โฆ Works on localhost โ your code and comments never leave your machine โฆ Surgical changes โ the CSS selector scopes your agent to the one element you commented on โฆ Multiple projects supported โ comments are routed to the right workspace by port โฆ One-command setup, one-command uninstall
Jun 13, 2026
rating_count is the Chrome Web Store ratings count, not a written-review count.
Media assets
Screenshots and videos on the listing.
Has promo video
Whether the listing includes at least one video.
Languages
Declared language locales.
Developer website
Listing exposes a developer website URL.
Contact email
Listing exposes a contact email.
Keyword in name
Case-insensitive substring match in the name.
Keyword in description
Case-insensitive substring match in the description.
Keyword occurrences in description
Count of case-insensitive occurrences in the description.
Category user-count percentile
Share of same-category extensions with fewer users (null if unknown).
These are transparent listing completeness / keyword signals, not a prediction of Chrome Web Store search ranking.