Skip to content
Back to skills

Visual Companion

ASecurity

Browser-based companion for showing mockups, diagrams, wireframes, and visual comparisons during design conversations. Uses chrome MCP tools. Falls back to text-only if unavailable. Use during brainstorming or any visual design work.

  • 3 stars
  • 0 votes
  • 0 copies
  • 3 views
  • Added May 28, 2026
ai-agentsgonode

Works with

  • terminal
  • cli
  • mcp

Security analysis

A100/100

Scanned May 28, 2026

npx -y skills add ArtemioPadilla/agent-triforce --skill visual-companion --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Visual Companion?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Visual Companion
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/artemiopadilla-visual-companion/badge)](https://www.skillsdirectory.com/skills/artemiopadilla-visual-companion)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: visual-companion
description: >
  Browser-based companion for showing mockups, diagrams, wireframes, and visual
  comparisons during design conversations. Uses chrome MCP tools. Falls back to
  text-only if unavailable. Use during brainstorming or any visual design work.
---

# Visual Brainstorming Companion

A browser-based tool for visual design work during: $ARGUMENTS

## Availability Check

Before offering the companion, check if browser tools are available:
- If `mcp__claude-in-chrome__*` tools are accessible: companion is available
- If not: proceed text-only, do not offer the companion, this is not an error

## Consent Flow

**Offer once per session, in its own message (no other content):**

> Some of what we're working on might be easier to explain if I can show it to you in a web browser. I can put together mockups, diagrams, comparisons, and other visuals as we go. Want to try it?

Wait for response. If declined, proceed text-only.

## Per-Question Decision Rule

Even after acceptance, decide FOR EACH QUESTION whether to use browser or terminal:

**Use the browser** when the user would understand better by SEEING it:
- Mockups and wireframes
- Layout comparisons (side-by-side)
- Architecture diagrams (boxes and arrows)
- Data flow visualizations
- State machine diagrams
- Color/typography comparisons

**Use the terminal** when the content is fundamentally TEXT:
- Requirements questions
- Conceptual choices
- Tradeoff lists
- Scope decisions
- Technical constraints
- A/B/C option descriptions

**A question about a UI topic is not automatically visual.** "What does personality mean here?" is terminal. "Which layout works better?" is browser.

## Rendering Guidelines

When rendering in the browser:

1. Create a new tab with `mcp__claude-in-chrome__tabs_create_mcp`
2. Use `mcp__claude-in-chrome__navigate` to load an HTML data URI or local file
3. Render with:
   - Tailwind CSS for styling (via CDN)
   - Clean, semantic HTML
   - Dark/light mode support
   - Responsive layout

### What to Render
- **Architecture diagrams**: HTML/CSS boxes with flexbox/grid, connecting lines
- **UI mockups**: Tailwind-styled components at realistic sizes
- **Comparisons**: Split-pane layouts, side-by-side options with labels
- **Data flows**: Labeled boxes with directional arrows
- **State machines**: Node-edge diagrams with transition labels

### Rendering Rules
- Keep each visual focused on ONE question or comparison
- Label everything clearly — the visual should be self-explanatory
- Use consistent colors: primary for focus, gray for context, red for warnings
- Include a title describing what the visual shows

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.